people_counting_from_images.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## people_counting_from_images\n",
    "\n",
    "In this notebook we will batch process a directory of images, printing the number of people counted in each image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "import json\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "images = list(Path(\"images/\").rglob(\"*.jpg\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[PosixPath('images/scenes/dog.jpg'),\n",
       " PosixPath('images/scenes/street.jpg'),\n",
       " PosixPath('images/scenes/farm.jpg'),\n",
       " PosixPath('images/scenes/pool.jpg'),\n",
       " PosixPath('images/scenes/supermarket.jpg'),\n",
       " PosixPath('images/scenes/crossing.jpg'),\n",
       " PosixPath('images/scenes/porsche.jpg'),\n",
       " PosixPath('images/scenes/people_car.jpg'),\n",
       " PosixPath('images/scenes/test-image3.jpg'),\n",
       " PosixPath('images/scenes/.ipynb_checkpoints/supermarket-checkpoint.jpg'),\n",
       " PosixPath('images/faces/me.jpg'),\n",
       " PosixPath('images/faces/idris.jpg'),\n",
       " PosixPath('images/faces/idris_2.jpg'),\n",
       " PosixPath('images/faces/family.jpg'),\n",
       " PosixPath('images/faces/face.jpg'),\n",
       " PosixPath('images/faces/couple.jpg'),\n",
       " PosixPath('images/faces/.ipynb_checkpoints/me-checkpoint.jpg')]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we can access the file name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'dog.jpg'"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "images[0].name"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "or get the path as a string"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'images/scenes/dog.jpg'"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "str(images[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Deepstack\n",
    "setup deepstack"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# deepstack credentials\n",
    "IP_ADDRESS = 'localhost'\n",
    "PORT = 80\n",
    "API_KEY =  \"\" # if you have not set an api_key, just pass an empty string"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import deepstack.core as ds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "dsobject = ds.DeepstackObject(IP_ADDRESS, PORT, API_KEY)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Loop over images getting all predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing image number 0 : images/scenes/dog.jpg\n",
      "Processing image number 1 : images/scenes/street.jpg\n",
      "Processing image number 2 : images/scenes/farm.jpg\n",
      "Processing image number 3 : images/scenes/pool.jpg\n",
      "Processing image number 4 : images/scenes/supermarket.jpg\n",
      "Processing image number 5 : images/scenes/crossing.jpg\n",
      "Processing image number 6 : images/scenes/porsche.jpg\n",
      "Processing image number 7 : images/scenes/people_car.jpg\n",
      "Processing image number 8 : images/scenes/test-image3.jpg\n",
      "Processing image number 9 : images/scenes/.ipynb_checkpoints/supermarket-checkpoint.jpg\n",
      "Processing image number 10 : images/faces/me.jpg\n",
      "Processing image number 11 : images/faces/idris.jpg\n",
      "Processing image number 12 : images/faces/idris_2.jpg\n",
      "Processing image number 13 : images/faces/family.jpg\n",
      "Processing image number 14 : images/faces/face.jpg\n",
      "Processing image number 15 : images/faces/couple.jpg\n",
      "Processing image number 16 : images/faces/.ipynb_checkpoints/me-checkpoint.jpg\n",
      "CPU times: user 86.6 ms, sys: 31.5 ms, total: 118 ms\n",
      "Wall time: 9.4 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "results = [] # keep predictions\n",
    "\n",
    "for i, img in enumerate(images):\n",
    "    print(f\"Processing image number {i} : {str(img)}\")\n",
    "    try:\n",
    "        with open(img, 'rb') as image_bytes:\n",
    "            predictions = dsobject.detect(image_bytes)\n",
    "            \n",
    "            results.append({\n",
    "                'file': str(img),\n",
    "                'predictions': predictions\n",
    "            })\n",
    "    except Exception as exc:\n",
    "        print(exc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The results are in a standardised format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'file': 'images/scenes/street.jpg',\n",
       " 'predictions': [{'confidence': 0.45400372,\n",
       "   'label': 'person',\n",
       "   'y_min': 537,\n",
       "   'x_min': 162,\n",
       "   'y_max': 790,\n",
       "   'x_max': 257},\n",
       "  {'confidence': 0.48486602,\n",
       "   'label': 'car',\n",
       "   'y_min': 168,\n",
       "   'x_min': 635,\n",
       "   'y_max': 291,\n",
       "   'x_max': 848},\n",
       "  {'confidence': 0.54467934,\n",
       "   'label': 'truck',\n",
       "   'y_min': 116,\n",
       "   'x_min': 888,\n",
       "   'y_max': 230,\n",
       "   'x_max': 1090},\n",
       "  {'confidence': 0.706636,\n",
       "   'label': 'car',\n",
       "   'y_min': 780,\n",
       "   'x_min': 2106,\n",
       "   'y_max': 1004,\n",
       "   'x_max': 2238},\n",
       "  {'confidence': 0.7713211,\n",
       "   'label': 'car',\n",
       "   'y_min': 550,\n",
       "   'x_min': 300,\n",
       "   'y_max': 774,\n",
       "   'x_max': 818},\n",
       "  {'confidence': 0.8798733,\n",
       "   'label': 'car',\n",
       "   'y_min': 402,\n",
       "   'x_min': 854,\n",
       "   'y_max': 586,\n",
       "   'x_max': 1289},\n",
       "  {'confidence': 0.91239434,\n",
       "   'label': 'car',\n",
       "   'y_min': 586,\n",
       "   'x_min': 1063,\n",
       "   'y_max': 957,\n",
       "   'x_max': 1784}]}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can write to json file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('outputs/results.json', 'w') as fp:\n",
    "    json.dump(results, fp)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets get the number of people in each image using a helper from deepstack-python"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "people = {}\n",
    "for pred in results:\n",
    "    people[Path(pred['file']).name] = len(ds.get_object_confidences(pred['predictions'], 'person'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'dog.jpg': 0,\n",
       " 'street.jpg': 1,\n",
       " 'farm.jpg': 0,\n",
       " 'pool.jpg': 3,\n",
       " 'supermarket.jpg': 0,\n",
       " 'crossing.jpg': 8,\n",
       " 'porsche.jpg': 0,\n",
       " 'people_car.jpg': 2,\n",
       " 'test-image3.jpg': 2,\n",
       " 'supermarket-checkpoint.jpg': 0,\n",
       " 'me.jpg': 1,\n",
       " 'idris.jpg': 1,\n",
       " 'idris_2.jpg': 1,\n",
       " 'family.jpg': 2,\n",
       " 'face.jpg': 4,\n",
       " 'couple.jpg': 2,\n",
       " 'me-checkpoint.jpg': 1}"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "people"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Visualise using pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.Series(people).plot.barh(figsize=(10,5)).set_xlabel('Person count');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}