VGG16_ImgPro.ipynb

{
  "cells": [
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "RbAFc0gCE1Yf"
      },
      "source": [
        "> ### **OPTIMAL DATA RANGE**\n",
        "---\n",
        "*   ClipLimit: 2-3\n",
        "*   Learning rate: 0.0001 < lr < 0.1\n",
        "*   Batch size: 32, 64, 128\n",
        "*   Dropout rate: 0,5 - 0,8\n",
        "*   Weight: auto"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "GS2cutmMFf7k"
      },
      "source": [
        "> ### **STRUCTURE**\n",
        "---\n",
        "\n",
        "*   Optimizer: Adam\n",
        "*   Activation function: ReLu (training), Softmax (final)\n",
        "*   Pooling: _Max Pooling_\n",
        "*   Kernel size: 130x242\n",
        "*   Training/testing/validation: 70/10/20\n",
        "*   Epoch: 30-65\n",
        "*   Class: 7\n",
        "*   Weight: ImageNet\n",
        "*   Loss: Cross-Entropy"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "RiQq7ALDHNRD"
      },
      "source": [
        " ## **DATA PREPARATION**\n",
        "---\n",
        "menyiapkan data (termasuk penerapan image enhancement)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "pip install -r \"Rupiah-PCR-Using-Image-Processing-and-CNN\\requirements.txt\""
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "rzc4-JxOI9Jj"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "import cv2\n",
        "import PIL\n",
        "from PIL import Image\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn import preprocessing, neighbors\n",
        "from sklearn.model_selection import train_test_split"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "izpw8VI-I9lt",
        "outputId": "9b0ba66a-3bef-4e3f-92cf-89f1426a0490"
      },
      "outputs": [],
      "source": [
        "path = r\"Rupiah-PCR-Using-Image-Processing-and-CNN\\datasetUang\"\n",
        "\n",
        "kelas = os.listdir(path)\n",
        "print(\"Classes list = \",kelas,\"\\n\\nNumber of classes = \", len(kelas))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "v4bmieGnJDvb"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "def imageEnhancement(image):\n",
        "  #eqhist biasa\n",
        "  gray_img_eqhist=cv2.equalizeHist(image)\n",
        "\n",
        "  #contrast Limited Adaptive Histogram Equalization\n",
        "  clahe=cv2.createCLAHE(clipLimit=2)\n",
        "  gray_img_clahe=clahe.apply(gray_img_eqhist)\n",
        "  ####img = cv2_imshow(gray_img_clahe)\n",
        "  return(gray_img_clahe)\n",
        "  #return(gray_img_eqhist)\n",
        "\n",
        "def imageSegmentation(image):\n",
        "  from skimage import filters\n",
        "  ret, otsu = cv2.threshold(image, 0, 255, cv2.THRESH_BINARY + \n",
        "                                            cv2.THRESH_OTSU)  \n",
        "  return(otsu)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "id": "s4SXYgp8JGRk",
        "outputId": "c7e628d6-8a20-4bcc-f967-5e915fbf232e"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "<PIL.Image.Image image mode=L size=552x246 at 0x19C735CECE0>\n",
            "(246, 552)\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=552x246>"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "img = cv2.imread(r\"Rupiah-PCR-Using-Image-Processing-and-CNN\\datasetUang\\50000\\50k_b_2022_14.png\")\n",
        "img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)\n",
        "x = imageEnhancement(img)\n",
        "x = imageSegmentation(x)\n",
        "\n",
        "z = PIL.Image.fromarray(x)\n",
        "print(z)\n",
        "\n",
        "print(img.shape)\n",
        "\n",
        "PIL.Image.fromarray(x)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "rlgOUif9Joap"
      },
      "outputs": [],
      "source": [
        "arrayNomer = os.listdir(path)\n",
        "dataset = \"datasetPreprocessing\"\n",
        "from os import listdir,makedirs\n",
        "from os.path import isfile,join\n",
        "\n",
        "def preprocessing(path,dspath,nilaike):\n",
        "  path = path #path\n",
        "  dstpath = dataset #folder tujuan\n",
        "  arrayNomer = os.listdir(path)\n",
        "  n=nilaike\n",
        "\n",
        "  try:\n",
        "      makedirs(dstpath+'/'+arrayNomer[n])\n",
        "  except:\n",
        "      print (\"Direktori sudah ada, citra akan di-replace dalam direktori yang sama\")\n",
        "\n",
        "  for dirs in path:\n",
        "    #print(os.listdir(path))\n",
        "    files = [f for f in listdir(path+'/'+arrayNomer[n]+'/') if isfile(join(path+'/'+arrayNomer[n]+'/',f))]\n",
        "    for image in files:\n",
        "      try:\n",
        "            img = cv2.imread(os.path.join(path+'/'+arrayNomer[n]+'/',image))\n",
        "            gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)\n",
        "            file = join(dstpath+'/'+arrayNomer[n]+'/',image)\n",
        "            enhance = imageEnhancement(gray)\n",
        "            otsu = imageSegmentation(enhance)\n",
        "            y = PIL.Image.fromarray(otsu)\n",
        "            y.save(file) #menyimpan file\n",
        "      except:\n",
        "            print (\"{} is not converted\".format(image))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nwksPw_tJ2wu",
        "outputId": "72e1b982-d7a9-4a6b-d3a0-03d21c66df8f"
      },
      "outputs": [],
      "source": [
        "for i in range (len(arrayNomer)):\n",
        "  print(\"Succesfully loaded class of\",arrayNomer[i])\n",
        "  preprocessing(path,dataset,i)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2BvypkIhJ9Pw",
        "outputId": "cd5a63a5-607d-4b11-a448-91a283842007"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "datasetPreprocessing/\n",
            "    1000/\n",
            "        1k_b_2016_1.png\n",
            "        1k_b_2016_10.png\n",
            "        1k_b_2016_11.png\n",
            "        1k_b_2016_12.png\n",
            "        1k_b_2016_13.png\n",
            "        1k_b_2016_14.png\n",
            "        1k_b_2016_15.png\n",
            "        1k_b_2016_16.png\n",
            "        1k_b_2016_17.png\n",
            "        1k_b_2016_18.png\n",
            "        1k_b_2016_19.png\n",
            "        1k_b_2016_2.png\n",
            "        1k_b_2016_20.png\n",
            "        1k_b_2016_21.png\n",
            "        1k_b_2016_22.png\n",
            "        1k_b_2016_23.png\n",
            "        1k_b_2016_24.png\n",
            "        1k_b_2016_25.png\n",
            "        1k_b_2016_3.png\n",
            "        1k_b_2016_4.png\n",
            "        1k_b_2016_5.png\n",
            "        1k_b_2016_6.png\n",
            "        1k_b_2016_7.png\n",
            "        1k_b_2016_8.png\n",
            "        1k_b_2016_9.png\n",
            "        1k_b_2022_1.png\n",
            "        1k_b_2022_10.png\n",
            "        1k_b_2022_11.png\n",
            "        1k_b_2022_12.png\n",
            "        1k_b_2022_13.png\n",
            "        1k_b_2022_14.png\n",
            "        1k_b_2022_15.png\n",
            "        1k_b_2022_16.png\n",
            "        1k_b_2022_17.png\n",
            "        1k_b_2022_18.png\n",
            "        1k_b_2022_19.png\n",
            "        1k_b_2022_2.png\n",
            "        1k_b_2022_20.png\n",
            "        1k_b_2022_21.png\n",
            "        1k_b_2022_22.png\n",
            "        1k_b_2022_23.png\n",
            "        1k_b_2022_24.png\n",
            "        1k_b_2022_25.png\n",
            "        1k_b_2022_3.png\n",
            "        1k_b_2022_4.png\n",
            "        1k_b_2022_5.png\n",
            "        1k_b_2022_6.png\n",
            "        1k_b_2022_7.png\n",
            "        1k_b_2022_8.png\n",
            "        1k_b_2022_9.png\n",
            "        1k_d_2016_1.png\n",
            "        1k_d_2016_10.png\n",
            "        1k_d_2016_11.png\n",
            "        1k_d_2016_12.png\n",
            "        1k_d_2016_13.png\n",
            "        1k_d_2016_14.png\n",
            "        1k_d_2016_15.png\n",
            "        1k_d_2016_16.png\n",
            "        1k_d_2016_17.png\n",
            "        1k_d_2016_18.png\n",
            "        1k_d_2016_19.png\n",
            "        1k_d_2016_2.png\n",
            "        1k_d_2016_20.png\n",
            "        1k_d_2016_21.png\n",
            "        1k_d_2016_22.png\n",
            "        1k_d_2016_23.png\n",
            "        1k_d_2016_24.png\n",
            "        1k_d_2016_25.png\n",
            "        1k_d_2016_3.png\n",
            "        1k_d_2016_4.png\n",
            "        1k_d_2016_5.png\n",
            "        1k_d_2016_6.png\n",
            "        1k_d_2016_7.png\n",
            "        1k_d_2016_8.png\n",
            "        1k_d_2016_9.png\n",
            "        1k_d_2022_1.png\n",
            "        1k_d_2022_10.png\n",
            "        1k_d_2022_11.png\n",
            "        1k_d_2022_12.png\n",
            "        1k_d_2022_13.png\n",
            "        1k_d_2022_14.png\n",
            "        1k_d_2022_15.png\n",
            "        1k_d_2022_16.png\n",
            "        1k_d_2022_17.png\n",
            "        1k_d_2022_18.png\n",
            "        1k_d_2022_19.png\n",
            "        1k_d_2022_2.png\n",
            "        1k_d_2022_20.png\n",
            "        1k_d_2022_21.png\n",
            "        1k_d_2022_22.png\n",
            "        1k_d_2022_23.png\n",
            "        1k_d_2022_24.png\n",
            "        1k_d_2022_25.png\n",
            "        1k_d_2022_3.png\n",
            "        1k_d_2022_4.png\n",
            "        1k_d_2022_5.png\n",
            "        1k_d_2022_6.png\n",
            "        1k_d_2022_7.png\n",
            "        1k_d_2022_8.png\n",
            "        1k_d_2022_9.png\n",
            "    10000/\n",
            "        10k_b_2016_1.png\n",
            "        10k_b_2016_10.png\n",
            "        10k_b_2016_11.png\n",
            "        10k_b_2016_12.png\n",
            "        10k_b_2016_13.png\n",
            "        10k_b_2016_14.png\n",
            "        10k_b_2016_15.png\n",
            "        10k_b_2016_16.png\n",
            "        10k_b_2016_17.png\n",
            "        10k_b_2016_18.png\n",
            "        10k_b_2016_19.png\n",
            "        10k_b_2016_2.png\n",
            "        10k_b_2016_20.png\n",
            "        10k_b_2016_21.png\n",
            "        10k_b_2016_22.png\n",
            "        10k_b_2016_23.png\n",
            "        10k_b_2016_24.png\n",
            "        10k_b_2016_25.png\n",
            "        10k_b_2016_3.png\n",
            "        10k_b_2016_4.png\n",
            "        10k_b_2016_5.png\n",
            "        10k_b_2016_6.png\n",
            "        10k_b_2016_7.png\n",
            "        10k_b_2016_8.png\n",
            "        10k_b_2016_9.png\n",
            "        10k_b_2022_1.png\n",
            "        10k_b_2022_10.png\n",
            "        10k_b_2022_11.png\n",
            "        10k_b_2022_12.png\n",
            "        10k_b_2022_13.png\n",
            "        10k_b_2022_14.png\n",
            "        10k_b_2022_15.png\n",
            "        10k_b_2022_16.png\n",
            "        10k_b_2022_17.png\n",
            "        10k_b_2022_18.png\n",
            "        10k_b_2022_19.png\n",
            "        10k_b_2022_2.png\n",
            "        10k_b_2022_20.png\n",
            "        10k_b_2022_21.png\n",
            "        10k_b_2022_22.png\n",
            "        10k_b_2022_23.png\n",
            "        10k_b_2022_24.png\n",
            "        10k_b_2022_25.png\n",
            "        10k_b_2022_3.png\n",
            "        10k_b_2022_4.png\n",
            "        10k_b_2022_5.png\n",
            "        10k_b_2022_6.png\n",
            "        10k_b_2022_7.png\n",
            "        10k_b_2022_8.png\n",
            "        10k_b_2022_9.png\n",
            "        10k_d_2016_1.png\n",
            "        10k_d_2016_10.png\n",
            "        10k_d_2016_11.png\n",
            "        10k_d_2016_12.png\n",
            "        10k_d_2016_13.png\n",
            "        10k_d_2016_14.png\n",
            "        10k_d_2016_15.png\n",
            "        10k_d_2016_16.png\n",
            "        10k_d_2016_17.png\n",
            "        10k_d_2016_18.png\n",
            "        10k_d_2016_19.png\n",
            "        10k_d_2016_2.png\n",
            "        10k_d_2016_20.png\n",
            "        10k_d_2016_21.png\n",
            "        10k_d_2016_22.png\n",
            "        10k_d_2016_23.png\n",
            "        10k_d_2016_24.png\n",
            "        10k_d_2016_25.png\n",
            "        10k_d_2016_3.png\n",
            "        10k_d_2016_4.png\n",
            "        10k_d_2016_5.png\n",
            "        10k_d_2016_6.png\n",
            "        10k_d_2016_7.png\n",
            "        10k_d_2016_8.png\n",
            "        10k_d_2016_9.png\n",
            "        10k_d_2022_1.png\n",
            "        10k_d_2022_10.png\n",
            "        10k_d_2022_11.png\n",
            "        10k_d_2022_12.png\n",
            "        10k_d_2022_13.png\n",
            "        10k_d_2022_14.png\n",
            "        10k_d_2022_15.png\n",
            "        10k_d_2022_16.png\n",
            "        10k_d_2022_17.png\n",
            "        10k_d_2022_18.png\n",
            "        10k_d_2022_19.png\n",
            "        10k_d_2022_2.png\n",
            "        10k_d_2022_20.png\n",
            "        10k_d_2022_21.png\n",
            "        10k_d_2022_22.png\n",
            "        10k_d_2022_23.png\n",
            "        10k_d_2022_24.png\n",
            "        10k_d_2022_25.png\n",
            "        10k_d_2022_3.png\n",
            "        10k_d_2022_4.png\n",
            "        10k_d_2022_5.png\n",
            "        10k_d_2022_6.png\n",
            "        10k_d_2022_7.png\n",
            "        10k_d_2022_8.png\n",
            "        10k_d_2022_9.png\n",
            "    100000/\n",
            "        100k_b_2016_1.png\n",
            "        100k_b_2016_10.png\n",
            "        100k_b_2016_11.png\n",
            "        100k_b_2016_12.png\n",
            "        100k_b_2016_13.png\n",
            "        100k_b_2016_14.png\n",
            "        100k_b_2016_15.png\n",
            "        100k_b_2016_16.png\n",
            "        100k_b_2016_17.png\n",
            "        100k_b_2016_18.png\n",
            "        100k_b_2016_19.png\n",
            "        100k_b_2016_2.png\n",
            "        100k_b_2016_20.png\n",
            "        100k_b_2016_21.png\n",
            "        100k_b_2016_22.png\n",
            "        100k_b_2016_23.png\n",
            "        100k_b_2016_24.png\n",
            "        100k_b_2016_25.png\n",
            "        100k_b_2016_3.png\n",
            "        100k_b_2016_4.png\n",
            "        100k_b_2016_5.png\n",
            "        100k_b_2016_6.png\n",
            "        100k_b_2016_7.png\n",
            "        100k_b_2016_8.png\n",
            "        100k_b_2016_9.png\n",
            "        100k_b_2022_1.png\n",
            "        100k_b_2022_10.png\n",
            "        100k_b_2022_11.png\n",
            "        100k_b_2022_12.png\n",
            "        100k_b_2022_13.png\n",
            "        100k_b_2022_14.png\n",
            "        100k_b_2022_15.png\n",
            "        100k_b_2022_16.png\n",
            "        100k_b_2022_17.png\n",
            "        100k_b_2022_18.png\n",
            "        100k_b_2022_19.png\n",
            "        100k_b_2022_2.png\n",
            "        100k_b_2022_20.png\n",
            "        100k_b_2022_21.png\n",
            "        100k_b_2022_22.png\n",
            "        100k_b_2022_23.png\n",
            "        100k_b_2022_24.png\n",
            "        100k_b_2022_25.png\n",
            "        100k_b_2022_3.png\n",
            "        100k_b_2022_4.png\n",
            "        100k_b_2022_5.png\n",
            "        100k_b_2022_6.png\n",
            "        100k_b_2022_7.png\n",
            "        100k_b_2022_8.png\n",
            "        100k_b_2022_9.png\n",
            "        100k_d_2016_1.png\n",
            "        100k_d_2016_10.png\n",
            "        100k_d_2016_11.png\n",
            "        100k_d_2016_12.png\n",
            "        100k_d_2016_13.png\n",
            "        100k_d_2016_14.png\n",
            "        100k_d_2016_15.png\n",
            "        100k_d_2016_16.png\n",
            "        100k_d_2016_17.png\n",
            "        100k_d_2016_18.png\n",
            "        100k_d_2016_19.png\n",
            "        100k_d_2016_2.png\n",
            "        100k_d_2016_20.png\n",
            "        100k_d_2016_21.png\n",
            "        100k_d_2016_22.png\n",
            "        100k_d_2016_23.png\n",
            "        100k_d_2016_24.png\n",
            "        100k_d_2016_25.png\n",
            "        100k_d_2016_3.png\n",
            "        100k_d_2016_4.png\n",
            "        100k_d_2016_5.png\n",
            "        100k_d_2016_6.png\n",
            "        100k_d_2016_7.png\n",
            "        100k_d_2016_8.png\n",
            "        100k_d_2016_9.png\n",
            "        100k_d_2022_1.png\n",
            "        100k_d_2022_10.png\n",
            "        100k_d_2022_11.png\n",
            "        100k_d_2022_12.png\n",
            "        100k_d_2022_13.png\n",
            "        100k_d_2022_14.png\n",
            "        100k_d_2022_15.png\n",
            "        100k_d_2022_16.png\n",
            "        100k_d_2022_17.png\n",
            "        100k_d_2022_18.png\n",
            "        100k_d_2022_19.png\n",
            "        100k_d_2022_2.png\n",
            "        100k_d_2022_20.png\n",
            "        100k_d_2022_21.png\n",
            "        100k_d_2022_22.png\n",
            "        100k_d_2022_23.png\n",
            "        100k_d_2022_24.png\n",
            "        100k_d_2022_25.png\n",
            "        100k_d_2022_3.png\n",
            "        100k_d_2022_4.png\n",
            "        100k_d_2022_5.png\n",
            "        100k_d_2022_6.png\n",
            "        100k_d_2022_7.png\n",
            "        100k_d_2022_8.png\n",
            "        100k_d_2022_9.png\n",
            "    2000/\n",
            "        2k_b_2016_1.png\n",
            "        2k_b_2016_10.png\n",
            "        2k_b_2016_11.png\n",
            "        2k_b_2016_12.png\n",
            "        2k_b_2016_13.png\n",
            "        2k_b_2016_14.png\n",
            "        2k_b_2016_15.png\n",
            "        2k_b_2016_16.png\n",
            "        2k_b_2016_17.png\n",
            "        2k_b_2016_18.png\n",
            "        2k_b_2016_19.png\n",
            "        2k_b_2016_2.png\n",
            "        2k_b_2016_20.png\n",
            "        2k_b_2016_21.png\n",
            "        2k_b_2016_22.png\n",
            "        2k_b_2016_23.png\n",
            "        2k_b_2016_24.png\n",
            "        2k_b_2016_25.png\n",
            "        2k_b_2016_3.png\n",
            "        2k_b_2016_4.png\n",
            "        2k_b_2016_5.png\n",
            "        2k_b_2016_6.png\n",
            "        2k_b_2016_7.png\n",
            "        2k_b_2016_8.png\n",
            "        2k_b_2016_9.png\n",
            "        2k_b_2022_1.png\n",
            "        2k_b_2022_10.png\n",
            "        2k_b_2022_11.png\n",
            "        2k_b_2022_12.png\n",
            "        2k_b_2022_13.png\n",
            "        2k_b_2022_14.png\n",
            "        2k_b_2022_15.png\n",
            "        2k_b_2022_16.png\n",
            "        2k_b_2022_17.png\n",
            "        2k_b_2022_18.png\n",
            "        2k_b_2022_19.png\n",
            "        2k_b_2022_2.png\n",
            "        2k_b_2022_20.png\n",
            "        2k_b_2022_21.png\n",
            "        2k_b_2022_22.png\n",
            "        2k_b_2022_23.png\n",
            "        2k_b_2022_24.png\n",
            "        2k_b_2022_25.png\n",
            "        2k_b_2022_3.png\n",
            "        2k_b_2022_4.png\n",
            "        2k_b_2022_5.png\n",
            "        2k_b_2022_6.png\n",
            "        2k_b_2022_7.png\n",
            "        2k_b_2022_8.png\n",
            "        2k_b_2022_9.png\n",
            "        2k_d_2016_1.png\n",
            "        2k_d_2016_10.png\n",
            "        2k_d_2016_11.png\n",
            "        2k_d_2016_12.png\n",
            "        2k_d_2016_13.png\n",
            "        2k_d_2016_14.png\n",
            "        2k_d_2016_15.png\n",
            "        2k_d_2016_16.png\n",
            "        2k_d_2016_17.png\n",
            "        2k_d_2016_18.png\n",
            "        2k_d_2016_19.png\n",
            "        2k_d_2016_2.png\n",
            "        2k_d_2016_20.png\n",
            "        2k_d_2016_21.png\n",
            "        2k_d_2016_22.png\n",
            "        2k_d_2016_23.png\n",
            "        2k_d_2016_24.png\n",
            "        2k_d_2016_25.png\n",
            "        2k_d_2016_3.png\n",
            "        2k_d_2016_4.png\n",
            "        2k_d_2016_5.png\n",
            "        2k_d_2016_6.png\n",
            "        2k_d_2016_7.png\n",
            "        2k_d_2016_8.png\n",
            "        2k_d_2016_9.png\n",
            "        2k_d_2022_1.png\n",
            "        2k_d_2022_10.png\n",
            "        2k_d_2022_11.png\n",
            "        2k_d_2022_12.png\n",
            "        2k_d_2022_13.png\n",
            "        2k_d_2022_14.png\n",
            "        2k_d_2022_15.png\n",
            "        2k_d_2022_16.png\n",
            "        2k_d_2022_17.png\n",
            "        2k_d_2022_18.png\n",
            "        2k_d_2022_19.png\n",
            "        2k_d_2022_2.png\n",
            "        2k_d_2022_20.png\n",
            "        2k_d_2022_21.png\n",
            "        2k_d_2022_22.png\n",
            "        2k_d_2022_23.png\n",
            "        2k_d_2022_24.png\n",
            "        2k_d_2022_25.png\n",
            "        2k_d_2022_3.png\n",
            "        2k_d_2022_4.png\n",
            "        2k_d_2022_5.png\n",
            "        2k_d_2022_6.png\n",
            "        2k_d_2022_7.png\n",
            "        2k_d_2022_8.png\n",
            "        2k_d_2022_9.png\n",
            "    20000/\n",
            "        20k_b_2016_1.png\n",
            "        20k_b_2016_10.png\n",
            "        20k_b_2016_11.png\n",
            "        20k_b_2016_12.png\n",
            "        20k_b_2016_13.png\n",
            "        20k_b_2016_14.png\n",
            "        20k_b_2016_15.png\n",
            "        20k_b_2016_16.png\n",
            "        20k_b_2016_17.png\n",
            "        20k_b_2016_18.png\n",
            "        20k_b_2016_19.png\n",
            "        20k_b_2016_2.png\n",
            "        20k_b_2016_20.png\n",
            "        20k_b_2016_21.png\n",
            "        20k_b_2016_22.png\n",
            "        20k_b_2016_23.png\n",
            "        20k_b_2016_24.png\n",
            "        20k_b_2016_25.png\n",
            "        20k_b_2016_3.png\n",
            "        20k_b_2016_4.png\n",
            "        20k_b_2016_5.png\n",
            "        20k_b_2016_6.png\n",
            "        20k_b_2016_7.png\n",
            "        20k_b_2016_8.png\n",
            "        20k_b_2016_9.png\n",
            "        20k_b_2022_1.png\n",
            "        20k_b_2022_10.png\n",
            "        20k_b_2022_11.png\n",
            "        20k_b_2022_12.png\n",
            "        20k_b_2022_13.png\n",
            "        20k_b_2022_14.png\n",
            "        20k_b_2022_15.png\n",
            "        20k_b_2022_16.png\n",
            "        20k_b_2022_17.png\n",
            "        20k_b_2022_18.png\n",
            "        20k_b_2022_19.png\n",
            "        20k_b_2022_2.png\n",
            "        20k_b_2022_20.png\n",
            "        20k_b_2022_21.png\n",
            "        20k_b_2022_22.png\n",
            "        20k_b_2022_23.png\n",
            "        20k_b_2022_24.png\n",
            "        20k_b_2022_25.png\n",
            "        20k_b_2022_3.png\n",
            "        20k_b_2022_4.png\n",
            "        20k_b_2022_5.png\n",
            "        20k_b_2022_6.png\n",
            "        20k_b_2022_7.png\n",
            "        20k_b_2022_8.png\n",
            "        20k_b_2022_9.png\n",
            "        20k_d_2016_1.png\n",
            "        20k_d_2016_10.png\n",
            "        20k_d_2016_11.png\n",
            "        20k_d_2016_12.png\n",
            "        20k_d_2016_13.png\n",
            "        20k_d_2016_14.png\n",
            "        20k_d_2016_15.png\n",
            "        20k_d_2016_16.png\n",
            "        20k_d_2016_17.png\n",
            "        20k_d_2016_18.png\n",
            "        20k_d_2016_19.png\n",
            "        20k_d_2016_2.png\n",
            "        20k_d_2016_20.png\n",
            "        20k_d_2016_21.png\n",
            "        20k_d_2016_22.png\n",
            "        20k_d_2016_23.png\n",
            "        20k_d_2016_24.png\n",
            "        20k_d_2016_25.png\n",
            "        20k_d_2016_3.png\n",
            "        20k_d_2016_4.png\n",
            "        20k_d_2016_5.png\n",
            "        20k_d_2016_6.png\n",
            "        20k_d_2016_7.png\n",
            "        20k_d_2016_8.png\n",
            "        20k_d_2016_9.png\n",
            "        20k_d_2022_1.png\n",
            "        20k_d_2022_10.png\n",
            "        20k_d_2022_11.png\n",
            "        20k_d_2022_12.png\n",
            "        20k_d_2022_13.png\n",
            "        20k_d_2022_14.png\n",
            "        20k_d_2022_15.png\n",
            "        20k_d_2022_16.png\n",
            "        20k_d_2022_17.png\n",
            "        20k_d_2022_18.png\n",
            "        20k_d_2022_19.png\n",
            "        20k_d_2022_2.png\n",
            "        20k_d_2022_20.png\n",
            "        20k_d_2022_21.png\n",
            "        20k_d_2022_22.png\n",
            "        20k_d_2022_23.png\n",
            "        20k_d_2022_24.png\n",
            "        20k_d_2022_25.png\n",
            "        20k_d_2022_3.png\n",
            "        20k_d_2022_4.png\n",
            "        20k_d_2022_5.png\n",
            "        20k_d_2022_6.png\n",
            "        20k_d_2022_7.png\n",
            "        20k_d_2022_8.png\n",
            "        20k_d_2022_9.png\n",
            "    5000/\n",
            "        5k_b_2016_1.png\n",
            "        5k_b_2016_10.png\n",
            "        5k_b_2016_11.png\n",
            "        5k_b_2016_12.png\n",
            "        5k_b_2016_13.png\n",
            "        5k_b_2016_14.png\n",
            "        5k_b_2016_15.png\n",
            "        5k_b_2016_16.png\n",
            "        5k_b_2016_17.png\n",
            "        5k_b_2016_18.png\n",
            "        5k_b_2016_19.png\n",
            "        5k_b_2016_2.png\n",
            "        5k_b_2016_20.png\n",
            "        5k_b_2016_21.png\n",
            "        5k_b_2016_22.png\n",
            "        5k_b_2016_23.png\n",
            "        5k_b_2016_24.png\n",
            "        5k_b_2016_25.png\n",
            "        5k_b_2016_3.png\n",
            "        5k_b_2016_4.png\n",
            "        5k_b_2016_5.png\n",
            "        5k_b_2016_6.png\n",
            "        5k_b_2016_7.png\n",
            "        5k_b_2016_8.png\n",
            "        5k_b_2016_9.png\n",
            "        5k_b_2022_1.png\n",
            "        5k_b_2022_10.png\n",
            "        5k_b_2022_11.png\n",
            "        5k_b_2022_12.png\n",
            "        5k_b_2022_13.png\n",
            "        5k_b_2022_14.png\n",
            "        5k_b_2022_15.png\n",
            "        5k_b_2022_16.png\n",
            "        5k_b_2022_17.png\n",
            "        5k_b_2022_18.png\n",
            "        5k_b_2022_19.png\n",
            "        5k_b_2022_2.png\n",
            "        5k_b_2022_20.png\n",
            "        5k_b_2022_21.png\n",
            "        5k_b_2022_22.png\n",
            "        5k_b_2022_23.png\n",
            "        5k_b_2022_24.png\n",
            "        5k_b_2022_25.png\n",
            "        5k_b_2022_3.png\n",
            "        5k_b_2022_4.png\n",
            "        5k_b_2022_5.png\n",
            "        5k_b_2022_6.png\n",
            "        5k_b_2022_7.png\n",
            "        5k_b_2022_8.png\n",
            "        5k_b_2022_9.png\n",
            "        5k_d_2016_1.png\n",
            "        5k_d_2016_10.png\n",
            "        5k_d_2016_11.png\n",
            "        5k_d_2016_12.png\n",
            "        5k_d_2016_13.png\n",
            "        5k_d_2016_14.png\n",
            "        5k_d_2016_15.png\n",
            "        5k_d_2016_16.png\n",
            "        5k_d_2016_17.png\n",
            "        5k_d_2016_18.png\n",
            "        5k_d_2016_19.png\n",
            "        5k_d_2016_2.png\n",
            "        5k_d_2016_20.png\n",
            "        5k_d_2016_21.png\n",
            "        5k_d_2016_22.png\n",
            "        5k_d_2016_23.png\n",
            "        5k_d_2016_24.png\n",
            "        5k_d_2016_25.png\n",
            "        5k_d_2016_3.png\n",
            "        5k_d_2016_4.png\n",
            "        5k_d_2016_5.png\n",
            "        5k_d_2016_6.png\n",
            "        5k_d_2016_7.png\n",
            "        5k_d_2016_8.png\n",
            "        5k_d_2016_9.png\n",
            "        5k_d_2022_1.png\n",
            "        5k_d_2022_10.png\n",
            "        5k_d_2022_11.png\n",
            "        5k_d_2022_12.png\n",
            "        5k_d_2022_13.png\n",
            "        5k_d_2022_14.png\n",
            "        5k_d_2022_15.png\n",
            "        5k_d_2022_16.png\n",
            "        5k_d_2022_17.png\n",
            "        5k_d_2022_18.png\n",
            "        5k_d_2022_19.png\n",
            "        5k_d_2022_2.png\n",
            "        5k_d_2022_20.png\n",
            "        5k_d_2022_21.png\n",
            "        5k_d_2022_22.png\n",
            "        5k_d_2022_23.png\n",
            "        5k_d_2022_24.png\n",
            "        5k_d_2022_25.png\n",
            "        5k_d_2022_3.png\n",
            "        5k_d_2022_4.png\n",
            "        5k_d_2022_5.png\n",
            "        5k_d_2022_6.png\n",
            "        5k_d_2022_7.png\n",
            "        5k_d_2022_8.png\n",
            "        5k_d_2022_9.png\n",
            "    50000/\n",
            "        50k_b_2016_1.png\n",
            "        50k_b_2016_10.png\n",
            "        50k_b_2016_11.png\n",
            "        50k_b_2016_12.png\n",
            "        50k_b_2016_13.png\n",
            "        50k_b_2016_14.png\n",
            "        50k_b_2016_15.png\n",
            "        50k_b_2016_16.png\n",
            "        50k_b_2016_17.png\n",
            "        50k_b_2016_18.png\n",
            "        50k_b_2016_19.png\n",
            "        50k_b_2016_2.png\n",
            "        50k_b_2016_20.png\n",
            "        50k_b_2016_21.png\n",
            "        50k_b_2016_22.png\n",
            "        50k_b_2016_23.png\n",
            "        50k_b_2016_24.png\n",
            "        50k_b_2016_25.png\n",
            "        50k_b_2016_3.png\n",
            "        50k_b_2016_4.png\n",
            "        50k_b_2016_5.png\n",
            "        50k_b_2016_6.png\n",
            "        50k_b_2016_7.png\n",
            "        50k_b_2016_8.png\n",
            "        50k_b_2016_9.png\n",
            "        50k_b_2022_1.png\n",
            "        50k_b_2022_10.png\n",
            "        50k_b_2022_11.png\n",
            "        50k_b_2022_12.png\n",
            "        50k_b_2022_13.png\n",
            "        50k_b_2022_14.png\n",
            "        50k_b_2022_15.png\n",
            "        50k_b_2022_16.png\n",
            "        50k_b_2022_17.png\n",
            "        50k_b_2022_18.png\n",
            "        50k_b_2022_19.png\n",
            "        50k_b_2022_2.png\n",
            "        50k_b_2022_20.png\n",
            "        50k_b_2022_21.png\n",
            "        50k_b_2022_22.png\n",
            "        50k_b_2022_23.png\n",
            "        50k_b_2022_24.png\n",
            "        50k_b_2022_25.png\n",
            "        50k_b_2022_3.png\n",
            "        50k_b_2022_4.png\n",
            "        50k_b_2022_5.png\n",
            "        50k_b_2022_6.png\n",
            "        50k_b_2022_7.png\n",
            "        50k_b_2022_8.png\n",
            "        50k_b_2022_9.png\n",
            "        50k_d_2016_1.png\n",
            "        50k_d_2016_10.png\n",
            "        50k_d_2016_11.png\n",
            "        50k_d_2016_12.png\n",
            "        50k_d_2016_13.png\n",
            "        50k_d_2016_14.png\n",
            "        50k_d_2016_15.png\n",
            "        50k_d_2016_16.png\n",
            "        50k_d_2016_17.png\n",
            "        50k_d_2016_18.png\n",
            "        50k_d_2016_19.png\n",
            "        50k_d_2016_2.png\n",
            "        50k_d_2016_20.png\n",
            "        50k_d_2016_21.png\n",
            "        50k_d_2016_22.png\n",
            "        50k_d_2016_23.png\n",
            "        50k_d_2016_24.png\n",
            "        50k_d_2016_25.png\n",
            "        50k_d_2016_3.png\n",
            "        50k_d_2016_4.png\n",
            "        50k_d_2016_5.png\n",
            "        50k_d_2016_6.png\n",
            "        50k_d_2016_7.png\n",
            "        50k_d_2016_8.png\n",
            "        50k_d_2016_9.png\n",
            "        50k_d_2022_1.png\n",
            "        50k_d_2022_10.png\n",
            "        50k_d_2022_11.png\n",
            "        50k_d_2022_12.png\n",
            "        50k_d_2022_13.png\n",
            "        50k_d_2022_14.png\n",
            "        50k_d_2022_15.png\n",
            "        50k_d_2022_16.png\n",
            "        50k_d_2022_17.png\n",
            "        50k_d_2022_18.png\n",
            "        50k_d_2022_19.png\n",
            "        50k_d_2022_2.png\n",
            "        50k_d_2022_20.png\n",
            "        50k_d_2022_21.png\n",
            "        50k_d_2022_22.png\n",
            "        50k_d_2022_23.png\n",
            "        50k_d_2022_24.png\n",
            "        50k_d_2022_25.png\n",
            "        50k_d_2022_3.png\n",
            "        50k_d_2022_4.png\n",
            "        50k_d_2022_5.png\n",
            "        50k_d_2022_6.png\n",
            "        50k_d_2022_7.png\n",
            "        50k_d_2022_8.png\n",
            "        50k_d_2022_9.png\n"
          ]
        }
      ],
      "source": [
        "import os\n",
        "\n",
        "dataset=\"datasetPreprocessing\"\n",
        "def daftar_file(startpath):\n",
        "    for root, dirs, files in os.walk(startpath):\n",
        "        level = root.replace(startpath, '').count(os.sep)\n",
        "        indent = ' ' * 4 * (level)\n",
        "        print('{}{}/'.format(indent, os.path.basename(root)))\n",
        "        subindent = ' ' * 4 * (level + 1)\n",
        "        for f in files:\n",
        "            print('{}{}'.format(subindent, f))\n",
        "\n",
        "daftar_file(dataset)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QB77nfSkKFB0",
        "outputId": "bd8872bb-8c53-4057-e2db-cebebfd8a196"
      },
      "outputs": [],
      "source": [
        "!pip install split-folders"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RdSJFwNsKYPU",
        "outputId": "01eda199-75c1-48c0-d593-8fb193175023"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "Copying files: 700 files [00:04, 142.82 files/s]\n"
          ]
        }
      ],
      "source": [
        "import splitfolders\n",
        "splitfolders.ratio(dataset, output=\"datasetSplit\", seed=1307, ratio=(0.7, 0.2, 0.1)) #70% : 10% : 20%"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hvw2D5W_LIuC",
        "outputId": "c9c4ce5e-8677-4860-d9b6-c9978b5941cd"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "datasetSplit/\n",
            "    test/\n",
            "        1000/\n",
            "            1k_b_2016_24.png\n",
            "            1k_b_2022_23.png\n",
            "            1k_b_2022_6.png\n",
            "            1k_b_2022_8.png\n",
            "            1k_d_2016_13.png\n",
            "            1k_d_2016_15.png\n",
            "            1k_d_2016_23.png\n",
            "            1k_d_2016_8.png\n",
            "            1k_d_2022_21.png\n",
            "            1k_d_2022_22.png\n",
            "        10000/\n",
            "            10k_b_2016_24.png\n",
            "            10k_b_2022_23.png\n",
            "            10k_b_2022_6.png\n",
            "            10k_b_2022_8.png\n",
            "            10k_d_2016_13.png\n",
            "            10k_d_2016_15.png\n",
            "            10k_d_2016_23.png\n",
            "            10k_d_2016_8.png\n",
            "            10k_d_2022_21.png\n",
            "            10k_d_2022_22.png\n",
            "        100000/\n",
            "            100k_b_2016_24.png\n",
            "            100k_b_2022_23.png\n",
            "            100k_b_2022_6.png\n",
            "            100k_b_2022_8.png\n",
            "            100k_d_2016_13.png\n",
            "            100k_d_2016_15.png\n",
            "            100k_d_2016_23.png\n",
            "            100k_d_2016_8.png\n",
            "            100k_d_2022_21.png\n",
            "            100k_d_2022_22.png\n",
            "        2000/\n",
            "            2k_b_2016_24.png\n",
            "            2k_b_2022_23.png\n",
            "            2k_b_2022_6.png\n",
            "            2k_b_2022_8.png\n",
            "            2k_d_2016_13.png\n",
            "            2k_d_2016_15.png\n",
            "            2k_d_2016_23.png\n",
            "            2k_d_2016_8.png\n",
            "            2k_d_2022_21.png\n",
            "            2k_d_2022_22.png\n",
            "        20000/\n",
            "            20k_b_2016_24.png\n",
            "            20k_b_2022_23.png\n",
            "            20k_b_2022_6.png\n",
            "            20k_b_2022_8.png\n",
            "            20k_d_2016_13.png\n",
            "            20k_d_2016_15.png\n",
            "            20k_d_2016_23.png\n",
            "            20k_d_2016_8.png\n",
            "            20k_d_2022_21.png\n",
            "            20k_d_2022_22.png\n",
            "        5000/\n",
            "            5k_b_2016_24.png\n",
            "            5k_b_2022_23.png\n",
            "            5k_b_2022_6.png\n",
            "            5k_b_2022_8.png\n",
            "            5k_d_2016_13.png\n",
            "            5k_d_2016_15.png\n",
            "            5k_d_2016_23.png\n",
            "            5k_d_2016_8.png\n",
            "            5k_d_2022_21.png\n",
            "            5k_d_2022_22.png\n",
            "        50000/\n",
            "            50k_b_2016_24.png\n",
            "            50k_b_2022_23.png\n",
            "            50k_b_2022_6.png\n",
            "            50k_b_2022_8.png\n",
            "            50k_d_2016_13.png\n",
            "            50k_d_2016_15.png\n",
            "            50k_d_2016_23.png\n",
            "            50k_d_2016_8.png\n",
            "            50k_d_2022_21.png\n",
            "            50k_d_2022_22.png\n",
            "    train/\n",
            "        1000/\n",
            "            1k_b_2016_10.png\n",
            "            1k_b_2016_11.png\n",
            "            1k_b_2016_12.png\n",
            "            1k_b_2016_14.png\n",
            "            1k_b_2016_17.png\n",
            "            1k_b_2016_18.png\n",
            "            1k_b_2016_19.png\n",
            "            1k_b_2016_2.png\n",
            "            1k_b_2016_20.png\n",
            "            1k_b_2016_21.png\n",
            "            1k_b_2016_22.png\n",
            "            1k_b_2016_23.png\n",
            "            1k_b_2016_25.png\n",
            "            1k_b_2016_3.png\n",
            "            1k_b_2016_4.png\n",
            "            1k_b_2016_5.png\n",
            "            1k_b_2016_6.png\n",
            "            1k_b_2016_8.png\n",
            "            1k_b_2016_9.png\n",
            "            1k_b_2022_1.png\n",
            "            1k_b_2022_10.png\n",
            "            1k_b_2022_13.png\n",
            "            1k_b_2022_14.png\n",
            "            1k_b_2022_15.png\n",
            "            1k_b_2022_17.png\n",
            "            1k_b_2022_18.png\n",
            "            1k_b_2022_20.png\n",
            "            1k_b_2022_21.png\n",
            "            1k_b_2022_22.png\n",
            "            1k_b_2022_24.png\n",
            "            1k_b_2022_3.png\n",
            "            1k_b_2022_4.png\n",
            "            1k_b_2022_5.png\n",
            "            1k_b_2022_7.png\n",
            "            1k_b_2022_9.png\n",
            "            1k_d_2016_1.png\n",
            "            1k_d_2016_10.png\n",
            "            1k_d_2016_11.png\n",
            "            1k_d_2016_12.png\n",
            "            1k_d_2016_16.png\n",
            "            1k_d_2016_17.png\n",
            "            1k_d_2016_18.png\n",
            "            1k_d_2016_19.png\n",
            "            1k_d_2016_20.png\n",
            "            1k_d_2016_21.png\n",
            "            1k_d_2016_22.png\n",
            "            1k_d_2016_24.png\n",
            "            1k_d_2016_25.png\n",
            "            1k_d_2016_3.png\n",
            "            1k_d_2016_4.png\n",
            "            1k_d_2016_5.png\n",
            "            1k_d_2016_6.png\n",
            "            1k_d_2016_7.png\n",
            "            1k_d_2022_1.png\n",
            "            1k_d_2022_10.png\n",
            "            1k_d_2022_12.png\n",
            "            1k_d_2022_13.png\n",
            "            1k_d_2022_15.png\n",
            "            1k_d_2022_16.png\n",
            "            1k_d_2022_18.png\n",
            "            1k_d_2022_19.png\n",
            "            1k_d_2022_20.png\n",
            "            1k_d_2022_23.png\n",
            "            1k_d_2022_24.png\n",
            "            1k_d_2022_25.png\n",
            "            1k_d_2022_5.png\n",
            "            1k_d_2022_6.png\n",
            "            1k_d_2022_7.png\n",
            "            1k_d_2022_8.png\n",
            "            1k_d_2022_9.png\n",
            "        10000/\n",
            "            10k_b_2016_10.png\n",
            "            10k_b_2016_11.png\n",
            "            10k_b_2016_12.png\n",
            "            10k_b_2016_14.png\n",
            "            10k_b_2016_17.png\n",
            "            10k_b_2016_18.png\n",
            "            10k_b_2016_19.png\n",
            "            10k_b_2016_2.png\n",
            "            10k_b_2016_20.png\n",
            "            10k_b_2016_21.png\n",
            "            10k_b_2016_22.png\n",
            "            10k_b_2016_23.png\n",
            "            10k_b_2016_25.png\n",
            "            10k_b_2016_3.png\n",
            "            10k_b_2016_4.png\n",
            "            10k_b_2016_5.png\n",
            "            10k_b_2016_6.png\n",
            "            10k_b_2016_8.png\n",
            "            10k_b_2016_9.png\n",
            "            10k_b_2022_1.png\n",
            "            10k_b_2022_10.png\n",
            "            10k_b_2022_13.png\n",
            "            10k_b_2022_14.png\n",
            "            10k_b_2022_15.png\n",
            "            10k_b_2022_17.png\n",
            "            10k_b_2022_18.png\n",
            "            10k_b_2022_20.png\n",
            "            10k_b_2022_21.png\n",
            "            10k_b_2022_22.png\n",
            "            10k_b_2022_24.png\n",
            "            10k_b_2022_3.png\n",
            "            10k_b_2022_4.png\n",
            "            10k_b_2022_5.png\n",
            "            10k_b_2022_7.png\n",
            "            10k_b_2022_9.png\n",
            "            10k_d_2016_1.png\n",
            "            10k_d_2016_10.png\n",
            "            10k_d_2016_11.png\n",
            "            10k_d_2016_12.png\n",
            "            10k_d_2016_16.png\n",
            "            10k_d_2016_17.png\n",
            "            10k_d_2016_18.png\n",
            "            10k_d_2016_19.png\n",
            "            10k_d_2016_20.png\n",
            "            10k_d_2016_21.png\n",
            "            10k_d_2016_22.png\n",
            "            10k_d_2016_24.png\n",
            "            10k_d_2016_25.png\n",
            "            10k_d_2016_3.png\n",
            "            10k_d_2016_4.png\n",
            "            10k_d_2016_5.png\n",
            "            10k_d_2016_6.png\n",
            "            10k_d_2016_7.png\n",
            "            10k_d_2022_1.png\n",
            "            10k_d_2022_10.png\n",
            "            10k_d_2022_12.png\n",
            "            10k_d_2022_13.png\n",
            "            10k_d_2022_15.png\n",
            "            10k_d_2022_16.png\n",
            "            10k_d_2022_18.png\n",
            "            10k_d_2022_19.png\n",
            "            10k_d_2022_20.png\n",
            "            10k_d_2022_23.png\n",
            "            10k_d_2022_24.png\n",
            "            10k_d_2022_25.png\n",
            "            10k_d_2022_5.png\n",
            "            10k_d_2022_6.png\n",
            "            10k_d_2022_7.png\n",
            "            10k_d_2022_8.png\n",
            "            10k_d_2022_9.png\n",
            "        100000/\n",
            "            100k_b_2016_10.png\n",
            "            100k_b_2016_11.png\n",
            "            100k_b_2016_12.png\n",
            "            100k_b_2016_14.png\n",
            "            100k_b_2016_17.png\n",
            "            100k_b_2016_18.png\n",
            "            100k_b_2016_19.png\n",
            "            100k_b_2016_2.png\n",
            "            100k_b_2016_20.png\n",
            "            100k_b_2016_21.png\n",
            "            100k_b_2016_22.png\n",
            "            100k_b_2016_23.png\n",
            "            100k_b_2016_25.png\n",
            "            100k_b_2016_3.png\n",
            "            100k_b_2016_4.png\n",
            "            100k_b_2016_5.png\n",
            "            100k_b_2016_6.png\n",
            "            100k_b_2016_8.png\n",
            "            100k_b_2016_9.png\n",
            "            100k_b_2022_1.png\n",
            "            100k_b_2022_10.png\n",
            "            100k_b_2022_13.png\n",
            "            100k_b_2022_14.png\n",
            "            100k_b_2022_15.png\n",
            "            100k_b_2022_17.png\n",
            "            100k_b_2022_18.png\n",
            "            100k_b_2022_20.png\n",
            "            100k_b_2022_21.png\n",
            "            100k_b_2022_22.png\n",
            "            100k_b_2022_24.png\n",
            "            100k_b_2022_3.png\n",
            "            100k_b_2022_4.png\n",
            "            100k_b_2022_5.png\n",
            "            100k_b_2022_7.png\n",
            "            100k_b_2022_9.png\n",
            "            100k_d_2016_1.png\n",
            "            100k_d_2016_10.png\n",
            "            100k_d_2016_11.png\n",
            "            100k_d_2016_12.png\n",
            "            100k_d_2016_16.png\n",
            "            100k_d_2016_17.png\n",
            "            100k_d_2016_18.png\n",
            "            100k_d_2016_19.png\n",
            "            100k_d_2016_20.png\n",
            "            100k_d_2016_21.png\n",
            "            100k_d_2016_22.png\n",
            "            100k_d_2016_24.png\n",
            "            100k_d_2016_25.png\n",
            "            100k_d_2016_3.png\n",
            "            100k_d_2016_4.png\n",
            "            100k_d_2016_5.png\n",
            "            100k_d_2016_6.png\n",
            "            100k_d_2016_7.png\n",
            "            100k_d_2022_1.png\n",
            "            100k_d_2022_10.png\n",
            "            100k_d_2022_12.png\n",
            "            100k_d_2022_13.png\n",
            "            100k_d_2022_15.png\n",
            "            100k_d_2022_16.png\n",
            "            100k_d_2022_18.png\n",
            "            100k_d_2022_19.png\n",
            "            100k_d_2022_20.png\n",
            "            100k_d_2022_23.png\n",
            "            100k_d_2022_24.png\n",
            "            100k_d_2022_25.png\n",
            "            100k_d_2022_5.png\n",
            "            100k_d_2022_6.png\n",
            "            100k_d_2022_7.png\n",
            "            100k_d_2022_8.png\n",
            "            100k_d_2022_9.png\n",
            "        2000/\n",
            "            2k_b_2016_10.png\n",
            "            2k_b_2016_11.png\n",
            "            2k_b_2016_12.png\n",
            "            2k_b_2016_14.png\n",
            "            2k_b_2016_17.png\n",
            "            2k_b_2016_18.png\n",
            "            2k_b_2016_19.png\n",
            "            2k_b_2016_2.png\n",
            "            2k_b_2016_20.png\n",
            "            2k_b_2016_21.png\n",
            "            2k_b_2016_22.png\n",
            "            2k_b_2016_23.png\n",
            "            2k_b_2016_25.png\n",
            "            2k_b_2016_3.png\n",
            "            2k_b_2016_4.png\n",
            "            2k_b_2016_5.png\n",
            "            2k_b_2016_6.png\n",
            "            2k_b_2016_8.png\n",
            "            2k_b_2016_9.png\n",
            "            2k_b_2022_1.png\n",
            "            2k_b_2022_10.png\n",
            "            2k_b_2022_13.png\n",
            "            2k_b_2022_14.png\n",
            "            2k_b_2022_15.png\n",
            "            2k_b_2022_17.png\n",
            "            2k_b_2022_18.png\n",
            "            2k_b_2022_20.png\n",
            "            2k_b_2022_21.png\n",
            "            2k_b_2022_22.png\n",
            "            2k_b_2022_24.png\n",
            "            2k_b_2022_3.png\n",
            "            2k_b_2022_4.png\n",
            "            2k_b_2022_5.png\n",
            "            2k_b_2022_7.png\n",
            "            2k_b_2022_9.png\n",
            "            2k_d_2016_1.png\n",
            "            2k_d_2016_10.png\n",
            "            2k_d_2016_11.png\n",
            "            2k_d_2016_12.png\n",
            "            2k_d_2016_16.png\n",
            "            2k_d_2016_17.png\n",
            "            2k_d_2016_18.png\n",
            "            2k_d_2016_19.png\n",
            "            2k_d_2016_20.png\n",
            "            2k_d_2016_21.png\n",
            "            2k_d_2016_22.png\n",
            "            2k_d_2016_24.png\n",
            "            2k_d_2016_25.png\n",
            "            2k_d_2016_3.png\n",
            "            2k_d_2016_4.png\n",
            "            2k_d_2016_5.png\n",
            "            2k_d_2016_6.png\n",
            "            2k_d_2016_7.png\n",
            "            2k_d_2022_1.png\n",
            "            2k_d_2022_10.png\n",
            "            2k_d_2022_12.png\n",
            "            2k_d_2022_13.png\n",
            "            2k_d_2022_15.png\n",
            "            2k_d_2022_16.png\n",
            "            2k_d_2022_18.png\n",
            "            2k_d_2022_19.png\n",
            "            2k_d_2022_20.png\n",
            "            2k_d_2022_23.png\n",
            "            2k_d_2022_24.png\n",
            "            2k_d_2022_25.png\n",
            "            2k_d_2022_5.png\n",
            "            2k_d_2022_6.png\n",
            "            2k_d_2022_7.png\n",
            "            2k_d_2022_8.png\n",
            "            2k_d_2022_9.png\n",
            "        20000/\n",
            "            20k_b_2016_10.png\n",
            "            20k_b_2016_11.png\n",
            "            20k_b_2016_12.png\n",
            "            20k_b_2016_14.png\n",
            "            20k_b_2016_17.png\n",
            "            20k_b_2016_18.png\n",
            "            20k_b_2016_19.png\n",
            "            20k_b_2016_2.png\n",
            "            20k_b_2016_20.png\n",
            "            20k_b_2016_21.png\n",
            "            20k_b_2016_22.png\n",
            "            20k_b_2016_23.png\n",
            "            20k_b_2016_25.png\n",
            "            20k_b_2016_3.png\n",
            "            20k_b_2016_4.png\n",
            "            20k_b_2016_5.png\n",
            "            20k_b_2016_6.png\n",
            "            20k_b_2016_8.png\n",
            "            20k_b_2016_9.png\n",
            "            20k_b_2022_1.png\n",
            "            20k_b_2022_10.png\n",
            "            20k_b_2022_13.png\n",
            "            20k_b_2022_14.png\n",
            "            20k_b_2022_15.png\n",
            "            20k_b_2022_17.png\n",
            "            20k_b_2022_18.png\n",
            "            20k_b_2022_20.png\n",
            "            20k_b_2022_21.png\n",
            "            20k_b_2022_22.png\n",
            "            20k_b_2022_24.png\n",
            "            20k_b_2022_3.png\n",
            "            20k_b_2022_4.png\n",
            "            20k_b_2022_5.png\n",
            "            20k_b_2022_7.png\n",
            "            20k_b_2022_9.png\n",
            "            20k_d_2016_1.png\n",
            "            20k_d_2016_10.png\n",
            "            20k_d_2016_11.png\n",
            "            20k_d_2016_12.png\n",
            "            20k_d_2016_16.png\n",
            "            20k_d_2016_17.png\n",
            "            20k_d_2016_18.png\n",
            "            20k_d_2016_19.png\n",
            "            20k_d_2016_20.png\n",
            "            20k_d_2016_21.png\n",
            "            20k_d_2016_22.png\n",
            "            20k_d_2016_24.png\n",
            "            20k_d_2016_25.png\n",
            "            20k_d_2016_3.png\n",
            "            20k_d_2016_4.png\n",
            "            20k_d_2016_5.png\n",
            "            20k_d_2016_6.png\n",
            "            20k_d_2016_7.png\n",
            "            20k_d_2022_1.png\n",
            "            20k_d_2022_10.png\n",
            "            20k_d_2022_12.png\n",
            "            20k_d_2022_13.png\n",
            "            20k_d_2022_15.png\n",
            "            20k_d_2022_16.png\n",
            "            20k_d_2022_18.png\n",
            "            20k_d_2022_19.png\n",
            "            20k_d_2022_20.png\n",
            "            20k_d_2022_23.png\n",
            "            20k_d_2022_24.png\n",
            "            20k_d_2022_25.png\n",
            "            20k_d_2022_5.png\n",
            "            20k_d_2022_6.png\n",
            "            20k_d_2022_7.png\n",
            "            20k_d_2022_8.png\n",
            "            20k_d_2022_9.png\n",
            "        5000/\n",
            "            5k_b_2016_10.png\n",
            "            5k_b_2016_11.png\n",
            "            5k_b_2016_12.png\n",
            "            5k_b_2016_14.png\n",
            "            5k_b_2016_17.png\n",
            "            5k_b_2016_18.png\n",
            "            5k_b_2016_19.png\n",
            "            5k_b_2016_2.png\n",
            "            5k_b_2016_20.png\n",
            "            5k_b_2016_21.png\n",
            "            5k_b_2016_22.png\n",
            "            5k_b_2016_23.png\n",
            "            5k_b_2016_25.png\n",
            "            5k_b_2016_3.png\n",
            "            5k_b_2016_4.png\n",
            "            5k_b_2016_5.png\n",
            "            5k_b_2016_6.png\n",
            "            5k_b_2016_8.png\n",
            "            5k_b_2016_9.png\n",
            "            5k_b_2022_1.png\n",
            "            5k_b_2022_10.png\n",
            "            5k_b_2022_13.png\n",
            "            5k_b_2022_14.png\n",
            "            5k_b_2022_15.png\n",
            "            5k_b_2022_17.png\n",
            "            5k_b_2022_18.png\n",
            "            5k_b_2022_20.png\n",
            "            5k_b_2022_21.png\n",
            "            5k_b_2022_22.png\n",
            "            5k_b_2022_24.png\n",
            "            5k_b_2022_3.png\n",
            "            5k_b_2022_4.png\n",
            "            5k_b_2022_5.png\n",
            "            5k_b_2022_7.png\n",
            "            5k_b_2022_9.png\n",
            "            5k_d_2016_1.png\n",
            "            5k_d_2016_10.png\n",
            "            5k_d_2016_11.png\n",
            "            5k_d_2016_12.png\n",
            "            5k_d_2016_16.png\n",
            "            5k_d_2016_17.png\n",
            "            5k_d_2016_18.png\n",
            "            5k_d_2016_19.png\n",
            "            5k_d_2016_20.png\n",
            "            5k_d_2016_21.png\n",
            "            5k_d_2016_22.png\n",
            "            5k_d_2016_24.png\n",
            "            5k_d_2016_25.png\n",
            "            5k_d_2016_3.png\n",
            "            5k_d_2016_4.png\n",
            "            5k_d_2016_5.png\n",
            "            5k_d_2016_6.png\n",
            "            5k_d_2016_7.png\n",
            "            5k_d_2022_1.png\n",
            "            5k_d_2022_10.png\n",
            "            5k_d_2022_12.png\n",
            "            5k_d_2022_13.png\n",
            "            5k_d_2022_15.png\n",
            "            5k_d_2022_16.png\n",
            "            5k_d_2022_18.png\n",
            "            5k_d_2022_19.png\n",
            "            5k_d_2022_20.png\n",
            "            5k_d_2022_23.png\n",
            "            5k_d_2022_24.png\n",
            "            5k_d_2022_25.png\n",
            "            5k_d_2022_5.png\n",
            "            5k_d_2022_6.png\n",
            "            5k_d_2022_7.png\n",
            "            5k_d_2022_8.png\n",
            "            5k_d_2022_9.png\n",
            "        50000/\n",
            "            50k_b_2016_10.png\n",
            "            50k_b_2016_11.png\n",
            "            50k_b_2016_12.png\n",
            "            50k_b_2016_14.png\n",
            "            50k_b_2016_17.png\n",
            "            50k_b_2016_18.png\n",
            "            50k_b_2016_19.png\n",
            "            50k_b_2016_2.png\n",
            "            50k_b_2016_20.png\n",
            "            50k_b_2016_21.png\n",
            "            50k_b_2016_22.png\n",
            "            50k_b_2016_23.png\n",
            "            50k_b_2016_25.png\n",
            "            50k_b_2016_3.png\n",
            "            50k_b_2016_4.png\n",
            "            50k_b_2016_5.png\n",
            "            50k_b_2016_6.png\n",
            "            50k_b_2016_8.png\n",
            "            50k_b_2016_9.png\n",
            "            50k_b_2022_1.png\n",
            "            50k_b_2022_10.png\n",
            "            50k_b_2022_13.png\n",
            "            50k_b_2022_14.png\n",
            "            50k_b_2022_15.png\n",
            "            50k_b_2022_17.png\n",
            "            50k_b_2022_18.png\n",
            "            50k_b_2022_20.png\n",
            "            50k_b_2022_21.png\n",
            "            50k_b_2022_22.png\n",
            "            50k_b_2022_24.png\n",
            "            50k_b_2022_3.png\n",
            "            50k_b_2022_4.png\n",
            "            50k_b_2022_5.png\n",
            "            50k_b_2022_7.png\n",
            "            50k_b_2022_9.png\n",
            "            50k_d_2016_1.png\n",
            "            50k_d_2016_10.png\n",
            "            50k_d_2016_11.png\n",
            "            50k_d_2016_12.png\n",
            "            50k_d_2016_16.png\n",
            "            50k_d_2016_17.png\n",
            "            50k_d_2016_18.png\n",
            "            50k_d_2016_19.png\n",
            "            50k_d_2016_20.png\n",
            "            50k_d_2016_21.png\n",
            "            50k_d_2016_22.png\n",
            "            50k_d_2016_24.png\n",
            "            50k_d_2016_25.png\n",
            "            50k_d_2016_3.png\n",
            "            50k_d_2016_4.png\n",
            "            50k_d_2016_5.png\n",
            "            50k_d_2016_6.png\n",
            "            50k_d_2016_7.png\n",
            "            50k_d_2022_1.png\n",
            "            50k_d_2022_10.png\n",
            "            50k_d_2022_12.png\n",
            "            50k_d_2022_13.png\n",
            "            50k_d_2022_15.png\n",
            "            50k_d_2022_16.png\n",
            "            50k_d_2022_18.png\n",
            "            50k_d_2022_19.png\n",
            "            50k_d_2022_20.png\n",
            "            50k_d_2022_23.png\n",
            "            50k_d_2022_24.png\n",
            "            50k_d_2022_25.png\n",
            "            50k_d_2022_5.png\n",
            "            50k_d_2022_6.png\n",
            "            50k_d_2022_7.png\n",
            "            50k_d_2022_8.png\n",
            "            50k_d_2022_9.png\n",
            "    val/\n",
            "        1000/\n",
            "            1k_b_2016_1.png\n",
            "            1k_b_2016_13.png\n",
            "            1k_b_2016_15.png\n",
            "            1k_b_2016_16.png\n",
            "            1k_b_2016_7.png\n",
            "            1k_b_2022_11.png\n",
            "            1k_b_2022_12.png\n",
            "            1k_b_2022_16.png\n",
            "            1k_b_2022_19.png\n",
            "            1k_b_2022_2.png\n",
            "            1k_b_2022_25.png\n",
            "            1k_d_2016_14.png\n",
            "            1k_d_2016_2.png\n",
            "            1k_d_2016_9.png\n",
            "            1k_d_2022_11.png\n",
            "            1k_d_2022_14.png\n",
            "            1k_d_2022_17.png\n",
            "            1k_d_2022_2.png\n",
            "            1k_d_2022_3.png\n",
            "            1k_d_2022_4.png\n",
            "        10000/\n",
            "            10k_b_2016_1.png\n",
            "            10k_b_2016_13.png\n",
            "            10k_b_2016_15.png\n",
            "            10k_b_2016_16.png\n",
            "            10k_b_2016_7.png\n",
            "            10k_b_2022_11.png\n",
            "            10k_b_2022_12.png\n",
            "            10k_b_2022_16.png\n",
            "            10k_b_2022_19.png\n",
            "            10k_b_2022_2.png\n",
            "            10k_b_2022_25.png\n",
            "            10k_d_2016_14.png\n",
            "            10k_d_2016_2.png\n",
            "            10k_d_2016_9.png\n",
            "            10k_d_2022_11.png\n",
            "            10k_d_2022_14.png\n",
            "            10k_d_2022_17.png\n",
            "            10k_d_2022_2.png\n",
            "            10k_d_2022_3.png\n",
            "            10k_d_2022_4.png\n",
            "        100000/\n",
            "            100k_b_2016_1.png\n",
            "            100k_b_2016_13.png\n",
            "            100k_b_2016_15.png\n",
            "            100k_b_2016_16.png\n",
            "            100k_b_2016_7.png\n",
            "            100k_b_2022_11.png\n",
            "            100k_b_2022_12.png\n",
            "            100k_b_2022_16.png\n",
            "            100k_b_2022_19.png\n",
            "            100k_b_2022_2.png\n",
            "            100k_b_2022_25.png\n",
            "            100k_d_2016_14.png\n",
            "            100k_d_2016_2.png\n",
            "            100k_d_2016_9.png\n",
            "            100k_d_2022_11.png\n",
            "            100k_d_2022_14.png\n",
            "            100k_d_2022_17.png\n",
            "            100k_d_2022_2.png\n",
            "            100k_d_2022_3.png\n",
            "            100k_d_2022_4.png\n",
            "        2000/\n",
            "            2k_b_2016_1.png\n",
            "            2k_b_2016_13.png\n",
            "            2k_b_2016_15.png\n",
            "            2k_b_2016_16.png\n",
            "            2k_b_2016_7.png\n",
            "            2k_b_2022_11.png\n",
            "            2k_b_2022_12.png\n",
            "            2k_b_2022_16.png\n",
            "            2k_b_2022_19.png\n",
            "            2k_b_2022_2.png\n",
            "            2k_b_2022_25.png\n",
            "            2k_d_2016_14.png\n",
            "            2k_d_2016_2.png\n",
            "            2k_d_2016_9.png\n",
            "            2k_d_2022_11.png\n",
            "            2k_d_2022_14.png\n",
            "            2k_d_2022_17.png\n",
            "            2k_d_2022_2.png\n",
            "            2k_d_2022_3.png\n",
            "            2k_d_2022_4.png\n",
            "        20000/\n",
            "            20k_b_2016_1.png\n",
            "            20k_b_2016_13.png\n",
            "            20k_b_2016_15.png\n",
            "            20k_b_2016_16.png\n",
            "            20k_b_2016_7.png\n",
            "            20k_b_2022_11.png\n",
            "            20k_b_2022_12.png\n",
            "            20k_b_2022_16.png\n",
            "            20k_b_2022_19.png\n",
            "            20k_b_2022_2.png\n",
            "            20k_b_2022_25.png\n",
            "            20k_d_2016_14.png\n",
            "            20k_d_2016_2.png\n",
            "            20k_d_2016_9.png\n",
            "            20k_d_2022_11.png\n",
            "            20k_d_2022_14.png\n",
            "            20k_d_2022_17.png\n",
            "            20k_d_2022_2.png\n",
            "            20k_d_2022_3.png\n",
            "            20k_d_2022_4.png\n",
            "        5000/\n",
            "            5k_b_2016_1.png\n",
            "            5k_b_2016_13.png\n",
            "            5k_b_2016_15.png\n",
            "            5k_b_2016_16.png\n",
            "            5k_b_2016_7.png\n",
            "            5k_b_2022_11.png\n",
            "            5k_b_2022_12.png\n",
            "            5k_b_2022_16.png\n",
            "            5k_b_2022_19.png\n",
            "            5k_b_2022_2.png\n",
            "            5k_b_2022_25.png\n",
            "            5k_d_2016_14.png\n",
            "            5k_d_2016_2.png\n",
            "            5k_d_2016_9.png\n",
            "            5k_d_2022_11.png\n",
            "            5k_d_2022_14.png\n",
            "            5k_d_2022_17.png\n",
            "            5k_d_2022_2.png\n",
            "            5k_d_2022_3.png\n",
            "            5k_d_2022_4.png\n",
            "        50000/\n",
            "            50k_b_2016_1.png\n",
            "            50k_b_2016_13.png\n",
            "            50k_b_2016_15.png\n",
            "            50k_b_2016_16.png\n",
            "            50k_b_2016_7.png\n",
            "            50k_b_2022_11.png\n",
            "            50k_b_2022_12.png\n",
            "            50k_b_2022_16.png\n",
            "            50k_b_2022_19.png\n",
            "            50k_b_2022_2.png\n",
            "            50k_b_2022_25.png\n",
            "            50k_d_2016_14.png\n",
            "            50k_d_2016_2.png\n",
            "            50k_d_2016_9.png\n",
            "            50k_d_2022_11.png\n",
            "            50k_d_2022_14.png\n",
            "            50k_d_2022_17.png\n",
            "            50k_d_2022_2.png\n",
            "            50k_d_2022_3.png\n",
            "            50k_d_2022_4.png\n"
          ]
        }
      ],
      "source": [
        "dataset = \"datasetSplit\"\n",
        "daftar_file(dataset)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "ToC7O1guLVMc"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import random\n",
        "from IPython.display import Image\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "\n",
        "from keras.utils import plot_model\n",
        "from sklearn.metrics import classification_report\n",
        "from collections import Counter\n",
        "import tensorflow as tf\n",
        "\n",
        "import keras\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from keras.callbacks import ReduceLROnPlateau\n",
        "from tensorflow.keras.applications.vgg16 import VGG16\n",
        "\n",
        "from keras import Model, layers\n",
        "from keras.models import Sequential\n",
        "from keras.optimizers import Adam, SGD\n",
        "from keras.layers import GlobalMaxPooling2D, GlobalAveragePooling2D, Dropout, Dense, Input, Conv2D, MaxPooling2D, Flatten,MaxPooling3D"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "25wCmnRGLWAB"
      },
      "outputs": [],
      "source": [
        "IMAGE_SIZE = (130, 242, 3)\n",
        "dataset=\"datasetSplit\"\n",
        "train_pred_test_folders = os.listdir(dataset)\n",
        "train_path = dataset+'/train'\n",
        "test_path = dataset+'/test'\n",
        "val_path = dataset+'/val'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 534
        },
        "id": "FiEY_6rHLYKz",
        "outputId": "9d4609ad-b9a0-4390-929a-2c655aa760b6"
      },
      "outputs": [],
      "source": [
        "quantity_train = {} \n",
        "quantity_test = {}\n",
        "quantity_val = {}\n",
        "for folder in os.listdir(train_path):\n",
        "    quantity_train[folder] = len(os.listdir(train_path+'/'+folder))\n",
        "\n",
        "for folder in os.listdir(test_path):\n",
        "    quantity_test[folder] = len(os.listdir(test_path+'/'+folder))\n",
        "\n",
        "for folder in os.listdir(val_path):\n",
        "    quantity_val[folder] = len(os.listdir(val_path+'/'+folder))\n",
        "\n",
        "quantity_train = pd.DataFrame(list(quantity_train.items()), index=range(0,len(quantity_train)), columns=['class','count'])\n",
        "quantity_test = pd.DataFrame(list(quantity_test.items()), index=range(0,len(quantity_test)), columns=['class','count'])\n",
        "quantity_val = pd.DataFrame(list(quantity_val.items()), index=range(0,len(quantity_val)), columns=['class','count'])\n",
        "\n",
        "figure, ax = plt.subplots(1,2,figsize=(20,5))\n",
        "sns.barplot(x='class',y='count',data=quantity_train,ax=ax[0])\n",
        "sns.barplot(x='class',y='count',data=quantity_test,ax=ax[1])\n",
        "\n",
        "print(\"Number of training set:\", sum(quantity_train['count'].values))\n",
        "print(\"Number of testing set:\",sum(quantity_test['count'].values))\n",
        "print(\"Number of prediction set:\",sum(quantity_val['count'].values))\n",
        "print(\"Total:\",sum(quantity_train['count'].values)+sum(quantity_test['count'].values)+sum(quantity_val['count'].values))\n",
        "\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "MGtlhk8QLaZh"
      },
      "outputs": [],
      "source": [
        "def save_history(i, history, model_name):\n",
        "    #convert the history.history dict to a pandas DataFrame:     \n",
        "    hist_df = pd.DataFrame(history.history) \n",
        "\n",
        "    # save ke json \n",
        "    hist_json_file = 'CNN_Model'+str(i)+'/'+model_name+'_history.json' \n",
        "    with open(hist_json_file, mode='w') as f:\n",
        "        hist_df.to_json(f)\n",
        "\n",
        "    # or save ke csv \n",
        "    hist_csv_file = 'CNN_Model'+str(i)+'/'+model_name+'_history.csv'\n",
        "    with open(hist_csv_file, mode='w') as f:\n",
        "        hist_df.to_csv(f)\n",
        "        \n",
        "def plot_accuracy_from_history(history, isinception=False):\n",
        "  try:\n",
        "    color = sns.color_palette()\n",
        "    if(isinception == False):\n",
        "        acc = history.history['acc']\n",
        "        val_acc = history.history['val_acc']\n",
        "    else:\n",
        "        acc = history.history['accuracy']\n",
        "        val_acc = history.history['val_accuracy']\n",
        "    \n",
        "\n",
        "    epochs = range(len(acc))\n",
        "\n",
        "    sns.lineplot(epochs, acc, label='Training Accuracy')\n",
        "    sns.lineplot(epochs, val_acc,label='Validation Accuracy')\n",
        "    plt.title('Training and Validation Accuracy')\n",
        "    plt.legend()\n",
        "    plt.figure()\n",
        "    plt.show()\n",
        "  except TypeError:\n",
        "    pass\n",
        "    \n",
        "def plot_loss_from_history(history):\n",
        "  try:\n",
        "    color = sns.color_palette()\n",
        "    loss = history.history['loss']\n",
        "    val_loss = history.history['val_loss']\n",
        "    \n",
        "    epochs = range(len(loss))\n",
        "    \n",
        "    sns.lineplot(epochs, loss,label='Training Loss')\n",
        "    sns.lineplot(epochs, val_loss, label='Validation Loss')\n",
        "    plt.title('Training and Validation Loss')\n",
        "    plt.legend()\n",
        "    plt.figure()\n",
        "    plt.show()\n",
        "  except TypeError:\n",
        "    pass\n",
        "    \n",
        "def do_history_stuff(i, history, history_file_name, isinception=False):\n",
        "    save_history(i, history, history_file_name)\n",
        "    plot_accuracy_from_history(history, isinception)\n",
        "    plot_loss_from_history(history)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "ie_x1Pw9LlHe"
      },
      "outputs": [],
      "source": [
        "def show_few_images(number_of_examples=2, predict_using_model=None):\n",
        "    figure1, ax1 = plt.subplots(number_of_examples,len(os.listdir(train_path)), figsize=(20,4*number_of_examples))\n",
        "    ax1 = ax1.reshape(-1)\n",
        "    axoff_fun = np.vectorize(lambda ax:ax.axis('off'))\n",
        "    axoff_fun(ax1)\n",
        "    axs = 0\n",
        "    for i, folder in enumerate(os.listdir(train_path)):\n",
        "        image_ids = os.listdir(os.path.join(train_path,folder))\n",
        "        for j in [random.randrange(0, len(image_ids)) for i in range(0,number_of_examples)]:\n",
        "            display = plt.imread(os.path.join(train_path,folder,image_ids[j]))\n",
        "            plt.axis('off')\n",
        "            ax1[axs].imshow(display)\n",
        "            title = 'True:'+folder\n",
        "            if(predict_using_model):\n",
        "                predicted_classname = inv_map_classes[np.argmax(vgg16_final_model.predict(np.array([display])))]\n",
        "                title = title+'\\nPredict :'+predicted_classname\n",
        "            ax1[axs].set_title(title)\n",
        "            axs=axs+1"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "xCuQKBilLnS9"
      },
      "outputs": [],
      "source": [
        "tf.keras.backend.clear_session()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "z8lwG1_kv0gx"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "def saveCNN_Model(i):\n",
        "  try: \n",
        "    os.mkdir(\"CNN_Model\"+str(i)) \n",
        "  except OSError as error: \n",
        "    print(error)\n",
        "  saveModel(i)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "4HbaR-kTv3TN"
      },
      "outputs": [],
      "source": [
        "def saveModel(i):\n",
        "  try: \n",
        "    os.mkdir(\"CNN_Model\"+str(i)+\"/model\") \n",
        "  except OSError as error: \n",
        "    print(error)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "TQdkty4ZLt0M"
      },
      "source": [
        " ## **VGG-16**\n",
        "---\n",
        "Visual Geometry Group (16 layers)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "jXvi5J4K3j9-"
      },
      "source": [
        "#### **<font color='Pink'>Data Generator</font>**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DgkzGvmbnDxg",
        "outputId": "bd49992d-14ec-4a71-81be-02045365a35c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Found 490 images belonging to 7 classes.\n",
            "Found 70 images belonging to 7 classes.\n",
            "Found 140 images belonging to 7 classes.\n"
          ]
        }
      ],
      "source": [
        "#normalisasi\n",
        "train_datagen = ImageDataGenerator(rescale = 1.0/255.,shear_range=0.2,zoom_range=0.2)\n",
        "train_generator16 = train_datagen.flow_from_directory(train_path,\n",
        "                                                    batch_size=32,\n",
        "                                                    shuffle=True,\n",
        "                                                    class_mode='categorical',\n",
        "                                                    target_size=(130, 242))\n",
        "\n",
        "validation_datagen = ImageDataGenerator(rescale = 1.0/255.)\n",
        "\n",
        "test_datagen = ImageDataGenerator(rescale = 1.0/255.,shear_range=0.2,zoom_range=0.2)\n",
        "test_generator16 = test_datagen.flow_from_directory(test_path, target_size=(130, 242),\n",
        "    batch_size=1,\n",
        "    shuffle=True,\n",
        "    class_mode='categorical')\n",
        "\n",
        "validation_generator16 = validation_datagen.flow_from_directory(val_path, shuffle=True, batch_size=1, class_mode='categorical', target_size=(130, 242))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Jr3pfWLTnF8_",
        "outputId": "a6b2bfd0-2659-44f6-8a42-f29101199cda"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "{0: '1000', 1: '10000', 2: '100000', 3: '2000', 4: '20000', 5: '5000', 6: '50000'}\n"
          ]
        }
      ],
      "source": [
        "inv_map_classes = {v: k for k, v in validation_generator16.class_indices.items()}\n",
        "#print(validation_generator16.class_indices)\n",
        "print(inv_map_classes)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "aActKL9O3mnY"
      },
      "source": [
        "#### **<font color='Pink'>Hyperparameter</font>**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "NjEEV1Mx385B"
      },
      "outputs": [],
      "source": [
        "import itertools\n",
        "import random\n",
        "\n",
        "array = [['ep1','ep2','ep3','ep4'], ['lr1', 'lr2', 'lr3', 'lr4'], [32, 64, 128], ['dr1', 'dr2', 'dr3', 'dr4']]\n",
        "\n",
        "parameter = []\n",
        "lr_values = array[1][1:]  \n",
        "\n",
        "for _ in range(1):\n",
        "    updated_array = [array[0], ['lr1'] + lr_values, array[2], array[3]]\n",
        "    parameter.extend(list(itertools.product(*updated_array)))\n",
        "\n",
        "parameter = [\n",
        "    [\n",
        "        random.randint(30, 38) if val == 'ep1' else\n",
        "        random.randint(39, 47) if val == 'ep2' else\n",
        "        random.randint(48, 56) if val == 'ep3' else\n",
        "        random.randint(57, 65) if val == 'ep4' else\n",
        "\n",
        "        random.uniform(0.0001, 0.0250249) if val == 'lr1' else\n",
        "        random.uniform(0.025025, 0.050049) if val == 'lr2' else\n",
        "        random.uniform(0.05005, 0.0750749) if val == 'lr3' else\n",
        "        random.uniform(0.075075, 0.1) if val == 'lr4' else\n",
        "\n",
        "        random.uniform(0.5,0.5749) if val == 'dr1' else\n",
        "        random.uniform(0.575,0.649) if val == 'dr2' else\n",
        "        random.uniform(0.65,0.7249) if val == 'dr3' else\n",
        "        random.uniform(0.725,0.8) if val == 'dr4' else\n",
        "\n",
        "        val\n",
        "        for val in combination\n",
        "    ]\n",
        "    for combination in parameter\n",
        "]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3r9T4AumF8JZ",
        "outputId": "f4f89d5d-c0c9-41c7-b0e3-743f28884496"
      },
      "outputs": [],
      "source": [
        "for combination in parameter:\n",
        "    print(combination)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "HMKIsY1N6XTH"
      },
      "source": [
        "#### **<font color='Pink'>Model Training</font>**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "oQwWy9EDvRfT"
      },
      "outputs": [],
      "source": [
        "from keras.layers import BatchNormalization\n",
        "def vgg16_training(i, vgg_epoch, learning_rate, batch_size, dropout_rate):\n",
        "  vgg16_model = VGG16(pooling='max', weights='imagenet', include_top=False, input_shape=(130,242,3), classes=7) #defaultnya AVERAGE POOLING\n",
        "  for layers in vgg16_model.layers:\n",
        "              layers.trainable=False\n",
        "\n",
        "  last_output_vgg16 = vgg16_model.layers[-1].output\n",
        "  vgg_16 = Flatten()(last_output_vgg16) #awalnya di bawah last output\n",
        "  vgg_16 = BatchNormalization()(vgg_16)\n",
        "  vgg_16 = Dense(128, activation = 'relu')(vgg_16)\n",
        "  vgg_16 = BatchNormalization()(vgg_16)\n",
        "  vgg_16 = Dropout(dropout_rate)(vgg_16)\n",
        "\n",
        "  #vgg_16 = Dense(64, activation = 'relu')(vgg_16)\n",
        "  #vgg_16 = Dropout(dropout_rate)(vgg_16)\n",
        "  vgg_16 = Dense(7, activation = 'softmax')(vgg_16)\n",
        "  vgg16_final_model = Model(vgg16_model.input, vgg_16)\n",
        "  \n",
        "  opt = Adam(learning_rate=learning_rate)\n",
        "  vgg16_final_model.compile(loss = 'categorical_crossentropy', optimizer= opt, metrics=['acc'])\n",
        "  #vgg16_final_model.summary()\n",
        "\n",
        "  plot_model(model=vgg16_final_model, show_shapes=True)\n",
        "\n",
        "  train_generator16.batch_size = batch_size\n",
        "\n",
        "  vgg16_filepath = 'CNN_Model'+str(i)+'/model/vgg_16_'+str(i)+'-saved-model-{epoch:02d}-acc-{val_acc:.2f}.hdf5'\n",
        "  vgg16_checkpoint = tf.keras.callbacks.ModelCheckpoint(vgg16_filepath, monitor='val_acc', verbose=1, save_best_only=True, mode='max')\n",
        "  early_stopping = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=5)\n",
        "  vgg16_history = vgg16_final_model.fit(train_generator16, epochs = vgg_epoch, validation_data = validation_generator16,callbacks=[vgg16_checkpoint,early_stopping],verbose=1)\n",
        "\n",
        "  do_history_stuff(i, vgg16_history, 'vgg16_model')\n",
        "  save_vgg16(i, vgg16_final_model)\n",
        "\n",
        "  predictionTest(i, vgg16_final_model)\n",
        "\n",
        "  return(vgg16_final_model)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "uzyGkyI39Dbv"
      },
      "outputs": [],
      "source": [
        "def save_vgg16(i, vgg16_final_model):\n",
        "  vgg16_final_model.save(('CNN_Model'+str(i)+'/vgg16-model'+str(i)+'.h5'))"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "TIFMfEwf8HR8"
      },
      "source": [
        "#### **<font color='Pink'>Evaluation</font>**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "siSMQNcU8oB-"
      },
      "outputs": [],
      "source": [
        "def acc_plot(vgg16_history):\n",
        "  plt.figure(figsize=(10, 8))\n",
        "  plt.plot(vgg16_history.history['acc'], label='train acc')\n",
        "  plt.plot(vgg16_history.history['val_acc'], label='val acc')\n",
        "  plt.legend()\n",
        "  plt.title('Accuracy')\n",
        "  plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "id": "E7alj36o8w5T"
      },
      "outputs": [],
      "source": [
        "def loss_plot(vgg16_history):\n",
        "  plt.figure(figsize=(10, 8))\n",
        "  plt.plot(vgg16_history.history['loss'], label='train loss')\n",
        "  plt.plot(vgg16_history.history['val_loss'], label='val loss')\n",
        "  plt.legend()\n",
        "  plt.title('Loss')\n",
        "  plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "hoxqXW5381w9"
      },
      "outputs": [],
      "source": [
        "def predictionTest(i, vgg16_final_model):\n",
        "  true_value = []\n",
        "  vgg_pred = []\n",
        "  for folder in os.listdir(test_path):\n",
        "      test_image_ids = os.listdir(os.path.join(test_path,folder))\n",
        "      \n",
        "      for image_id in test_image_ids[:int(len(test_image_ids))]:\n",
        "          path = os.path.join(test_path,folder,image_id)\n",
        "          #print(path)\n",
        "          true_value.append(test_generator16.class_indices[folder])\n",
        "          img = cv2.resize(cv2.imread(path),(242,130))\n",
        "          img_normalized = img/255\n",
        "          #vgg\n",
        "          vgg16_image_prediction = np.argmax(vgg16_final_model.predict(np.array([img_normalized]), verbose = 0)) #verbose biar gak ngeprint\n",
        "          vgg_pred.append(vgg16_image_prediction)\n",
        "\n",
        "  print(\"\\n\")\n",
        "  clf_report(i, true_value, vgg_pred)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "pJV_qqpT86bD"
      },
      "outputs": [],
      "source": [
        "arr_accuracy16 = []\n",
        "from sklearn.metrics import confusion_matrix\n",
        "import itertools\n",
        "#from mlxtend.plotting import plot_confusion_matrix\n",
        "def clf_report(k, true_value, model_pred):\n",
        "    classes = validation_generator16.class_indices.keys()\n",
        "    TP_count = [true_value[i] == model_pred[i] for i in range(len(true_value))]\n",
        "    model_accuracy = np.sum(TP_count)/len(TP_count)\n",
        "    print('Model Accuracy', model_accuracy)\n",
        "    arr_accuracy16.append(model_accuracy)\n",
        "    plt.figure(figsize=(7,7))\n",
        "    cm = confusion_matrix(true_value,model_pred)\n",
        "    plt.imshow(cm,interpolation='nearest',cmap=plt.cm.viridis)\n",
        "    plt.title('Confusion Matrix')\n",
        "    plt.colorbar()\n",
        "    tick_marks = np.arange(len(classes))\n",
        "    plt.xticks(tick_marks, classes, rotation=45)\n",
        "    plt.yticks(tick_marks, classes)\n",
        "    thresh = cm.max()*0.8\n",
        "    for i,j in itertools.product(range(cm.shape[0]),range(cm.shape[1])):\n",
        "        plt.text(j,i,cm[i,j],\n",
        "                horizontalalignment=\"center\",\n",
        "                color=\"black\" if cm[i,j] > thresh else \"white\")\n",
        "        pass\n",
        "    \n",
        "    plt.ylabel('True Label')\n",
        "    plt.xlabel('Predicted Label')\n",
        "    pass\n",
        "    \n",
        "    plt.savefig(\"CNN_Model\"+str(k)+\"/conf_matrix\"+str(k)+\".png\")\n",
        "    print(classification_report(true_value, model_pred, target_names = list(classes)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "dict_keys(['1000', '10000', '100000', '2000', '20000', '5000', '50000'])\n"
          ]
        }
      ],
      "source": [
        "print(validation_generator16.class_indices.keys())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Empty DataFrame\n",
            "Columns: [Epoch, Learning Rate, Batch Size, Dropout Rate, Accuracy]\n",
            "Index: []\n"
          ]
        }
      ],
      "source": [
        "resultTable = pd.DataFrame(columns=['Epoch', 'Learning Rate', 'Batch Size', 'Dropout Rate','Accuracy'])\n",
        "print(resultTable)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "nr5k_MjPDmTN"
      },
      "source": [
        "#### **<font color='Pink'>Code Running</font>**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "j = 0\n",
        "for i in range (j,len(parameter)):\n",
        "  vgg_epoch = (parameter[i][0])\n",
        "  learning_rate = (parameter[i][1])\n",
        "  batch_size = (parameter[i][2])\n",
        "  dropout_rate = (parameter[i][3])\n",
        "  i=i+1\n",
        "  saveCNN_Model(i)\n",
        "  print(\"CNN_Model\",i,\"↓\")\n",
        "  print(\"HYPERPARAMETER\".center(100,\"─\"))\n",
        "  print(\"vgg epoch:\",vgg_epoch)\n",
        "  print(\"learning rate:\",learning_rate)\n",
        "  print(\"batch size:\",batch_size)\n",
        "  print(\"dropout rate:\",dropout_rate)\n",
        "  print(\"\".center(100,\"─\"))\n",
        "  vgg16_training(i, vgg_epoch, learning_rate, batch_size, dropout_rate)\n",
        "  new_row = {'Epoch': vgg_epoch, 'Learning Rate': learning_rate, 'Batch Size': batch_size, 'Dropout Rate': dropout_rate, 'Accuracy': arr_accuracy16[-1]}\n",
        "  resultTable = resultTable.append(new_row, ignore_index=True)\n",
        "  resultTable.index = resultTable.index + (j+1)\n",
        "  resultTable.to_excel(\"resultTable.xlsx\")\n",
        "  print(\"\".center(100,\"─\"))"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.9"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}