Pointcloud_LaneMarking_Detection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Pointcloud Visualization\n",
    "\n",
    "##### Idea\n",
    "\n",
    "To visualize the given LIDAR pointcloud data we first needed to convert the latitude and longitude information to a cartesian coordinate system that has X, Y and Z values to specify a single point. We can do this by using the latitude and longitude information to calculate the UTM coordinates for each point. As a result we will get a value for the easting (X) and northing (Y). The altitude value, which is already given as meters can thus be seen as the Z value in our new coordinate system. After saving the data to a new file we can then use http://lidarview.com/ to visualize the pointcloud to get a first impression of the data.\n",
    "\n",
    "##### Problems\n",
    "\n",
    "Our initial thought was that the intesity of each point in our point cloud is very important to determine the the lane markings in the point cloud. In fact, the lane marking are on a flat surface and the marking itself can also be seen as a flat line on the flat surface. Therfore, we cannot use the shape or variance in the point cloud to detect the lane markings. However, the intensity of each LIDAR point should contain enough information to recover the lane marking from the flat surface since the lane marking should have a higher intensity compared to the plain asphalt or concrete of the street. The problem during the visualization was that the visualization tool that we were using could not visualize the intensity of each point. As you can see in the following image, the visualization uses different colors to denote the elevation instead of the intensity of each point. Although the tool had an option for choosing the intensity the visualization did not work, maybe because the intensity was in a wrong format. Finally, we tried to use this tool https://asmaloney.com/software/ which was able to visualize the intensity in a proper manner. You can now easily recognize the lane markings which shows, that we can use the intensity to identify the lane markings.\n",
    "\n",
    "##### Results\n",
    "![Full LIDAR Pointcloud Visualization](FullView.png \"Full LIDAR Pointcloud View\")\n",
    "\n",
    "![Intensity LIDAR Pointcloud Visualization](IntensityView.png \"Intensity LIDAR Pointcloud View\")\n",
    "\n",
    "###### Code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import geopandas as gp\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from shapely import wkt\n",
    "from sklearn.cluster import DBSCAN\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from scipy.spatial.distance import pdist\n",
    "from scipy.spatial.distance import squareform\n",
    "\n",
    "import utm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_fuse_to_df(df_name):\n",
    "    data = []\n",
    "    \n",
    "    with open(df_name) as f:  \n",
    "        line = f.readline()\n",
    "        while line:\n",
    "            d = line.split()\n",
    "            data.append(d)\n",
    "            line = f.readline()\n",
    "\n",
    "    a = np.array(data)\n",
    "    \n",
    "    pointcloud_df = pd.DataFrame()\n",
    "    pointcloud_df[\"Latitude\"] = a[:,0]\n",
    "    pointcloud_df[\"Longitude\"] = a[:,1]\n",
    "    pointcloud_df[\"Altitude\"] = a[:,2]\n",
    "    pointcloud_df[\"Intensity\"] = a[:,3]\n",
    "\n",
    "    return pointcloud_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_fuse(pointcloud_df, min_x = 0.0, min_y = 0.0, min_z = 0.0):\n",
    "    pointcloud_df[\"Latitude\"] = pd.to_numeric(pointcloud_df[\"Latitude\"])\n",
    "    pointcloud_df[\"Longitude\"] = pd.to_numeric(pointcloud_df[\"Longitude\"])\n",
    "    pointcloud_df[\"Altitude\"] = pd.to_numeric(pointcloud_df[\"Altitude\"])\n",
    "    pointcloud_df[\"Intensity\"] = pd.to_numeric(pointcloud_df[\"Intensity\"])\n",
    "    pointcloud_df[\"Easting\"] = pointcloud_df.apply(lambda x: utm.from_latlon(x[\"Latitude\"], x[\"Longitude\"])[0], axis = 1)\n",
    "    pointcloud_df[\"Northing\"] = pointcloud_df.apply(lambda x: utm.from_latlon(x[\"Latitude\"], x[\"Longitude\"])[1], axis = 1)\n",
    "    \n",
    "    if min_x == 0:\n",
    "        min_x = pointcloud_df[\"Easting\"].min()\n",
    "    if min_y == 0:\n",
    "        min_y = pointcloud_df[\"Northing\"].min()   \n",
    "    if min_z == 0:\n",
    "        min_z = pointcloud_df[\"Altitude\"].min()    \n",
    "        \n",
    "    utm_coords = utm.from_latlon(pointcloud_df.loc[0,\"Latitude\"], pointcloud_df.loc[0,\"Longitude\"])\n",
    "\n",
    "    zone_number = utm_coords[2]\n",
    "    zone_letter = utm_coords[3]\n",
    "        \n",
    "    pointcloud_df[\"Easting\"] = pointcloud_df[\"Easting\"] - min_x    \n",
    "    pointcloud_df[\"Northing\"] = pointcloud_df[\"Northing\"] - min_y\n",
    "    pointcloud_df[\"Altitude\"] = pointcloud_df[\"Altitude\"] - min_z\n",
    "    \n",
    "    return pointcloud_df, (min_x, min_y, min_z), (zone_number, zone_letter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Latitude</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>Altitude</th>\n",
       "      <th>Intensity</th>\n",
       "      <th>Easting</th>\n",
       "      <th>Northing</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>45.903883</td>\n",
       "      <td>11.028414</td>\n",
       "      <td>10.6604</td>\n",
       "      <td>10</td>\n",
       "      <td>108.226321</td>\n",
       "      <td>65.043547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45.903683</td>\n",
       "      <td>11.028221</td>\n",
       "      <td>12.6662</td>\n",
       "      <td>5</td>\n",
       "      <td>93.822923</td>\n",
       "      <td>42.445590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>45.903683</td>\n",
       "      <td>11.028222</td>\n",
       "      <td>12.6415</td>\n",
       "      <td>7</td>\n",
       "      <td>93.938542</td>\n",
       "      <td>42.355143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>45.903682</td>\n",
       "      <td>11.028224</td>\n",
       "      <td>12.6263</td>\n",
       "      <td>7</td>\n",
       "      <td>94.080686</td>\n",
       "      <td>42.289830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>45.903681</td>\n",
       "      <td>11.028226</td>\n",
       "      <td>12.6188</td>\n",
       "      <td>7</td>\n",
       "      <td>94.233803</td>\n",
       "      <td>42.220349</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Latitude  Longitude  Altitude  Intensity     Easting   Northing\n",
       "0  45.903883  11.028414   10.6604         10  108.226321  65.043547\n",
       "1  45.903683  11.028221   12.6662          5   93.822923  42.445590\n",
       "2  45.903683  11.028222   12.6415          7   93.938542  42.355143\n",
       "3  45.903682  11.028224   12.6263          7   94.080686  42.289830\n",
       "4  45.903681  11.028226   12.6188          7   94.233803  42.220349"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pointcloud_df = read_fuse_to_df('./data/pointcloud.fuse')\n",
    "pointcloud_df, (min_x, min_y, min_z), (number, letter) = convert_fuse(pointcloud_df)\n",
    "\n",
    "pointcloud_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Easting</th>\n",
       "      <th>Northing</th>\n",
       "      <th>Altitude</th>\n",
       "      <th>Intensity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>108.226321</td>\n",
       "      <td>65.043547</td>\n",
       "      <td>10.6604</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>93.822923</td>\n",
       "      <td>42.445590</td>\n",
       "      <td>12.6662</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>93.938542</td>\n",
       "      <td>42.355143</td>\n",
       "      <td>12.6415</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>94.080686</td>\n",
       "      <td>42.289830</td>\n",
       "      <td>12.6263</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>94.233803</td>\n",
       "      <td>42.220349</td>\n",
       "      <td>12.6188</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Easting   Northing  Altitude  Intensity\n",
       "0  108.226321  65.043547   10.6604         10\n",
       "1   93.822923  42.445590   12.6662          5\n",
       "2   93.938542  42.355143   12.6415          7\n",
       "3   94.080686  42.289830   12.6263          7\n",
       "4   94.233803  42.220349   12.6188          7"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xyzi_df = pointcloud_df[[\"Easting\", \"Northing\", \"Altitude\", \"Intensity\"]]\n",
    "xyzi_df.to_csv(\"./data/pointcloud.xyz\", sep=\" \", header=False, index=False)\n",
    "xyzi_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Pointcloud Filtering\n",
    "\n",
    "##### Idea\n",
    "To identify the lanes in a much better way we first tried to remove all the noise from the data. In our case that means that we wanted to remove points with an intensity that is too high or too low to be a candidate lane marking point. Since different light conditions throughout the day might influence the intensity value, we first calculated the mean and standard deviation intensity for all points. The assumption here is that the lane marking points have a higher intensity value compared to the mean of all pointcloud points. We could have also set an adaptive threshold per pointcloud slice. However, since there was no slice information left we had to use this approach. The nefit of an adaptive threshold would have been that this approach is more robust against different light conditions within the current scene (e.g. a car that is just about to enter a tunnel but has not entered the tunnel yet). The first figure in the results section below shows the lane marking cloud point after the first filtering process.\n",
    "\n",
    "As you can also see in the figure, there are a lot of points off the road that obviosly do not belong to the lanes. To reduce the number of points that are off the road we used the actual trajectory information in order to remove points that are too far away from the actual trajectory line. We use a threshold of 20 meters to remove all points that are further away. That way we can reduce the number of points from 32191 to 18496. The second figure in the results below shows the remaining points after filtering based on the trajectory lines.\n",
    "\n",
    "##### Problems\n",
    "One major problem was that we could not use an adaptive threshold to remove more noise. While we could have computed an adaptive threshold in the 3 dimensional space, this approach would have been too computationaly expensive especially for the large dataset that we had in the beginning. A problem that might occur with our current algorithm setup are the boundaries for our threshold. Right now we tried to tweek the threshold boundaries to get good results. It would be good to test our current thresholds on larger datasets to see if our approach is valid or if we really need to use an adaptive threshold.\n",
    "\n",
    "##### Results\n",
    "![First LIDAR Pointcloud Filter](FirstFilter.png \"First LIDAR Pointcloud Filter\")\n",
    "![Trajectory LIDAR Pointcloud Filter](TrajectoryFilter.png \"Trajectory LIDAR Pointcloud Filter\")\n",
    "\n",
    "##### Code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_to_shape(row):\n",
    "    return wkt.loads(\"POINT (\" + str(row[\"Easting\"]) + \" \" + str(row[\"Northing\"]) +\" \" + str(row[\"Altitude\"]) + \")\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "def filter_by_mean_value(pointcloud_df, left_bound, right_bound):\n",
    "\n",
    "    mean = pointcloud_df[\"Intensity\"].mean()\n",
    "    std = pointcloud_df[\"Intensity\"].std()\n",
    "\n",
    "    lanes_df = pointcloud_df[pointcloud_df[\"Intensity\"] > mean + left_bound * std]\n",
    "    lanes_df = lanes_df[lanes_df[\"Intensity\"] < mean + right_bound * std ]\n",
    "\n",
    "    print(\"MEAN FILTER:\")\n",
    "    print(\"============\")\n",
    "    print(\"Intensity - Mean value:      \", mean)\n",
    "    print(\"Intensity - Std value:       \", std)\n",
    "    print(\"Intensity - Lower bound:     \", mean + 1 * std)\n",
    "    print(\"Intensity - Upper bound:     \", mean + 7 * std)\n",
    "    print(\"Intensity - Filtered points: \", len(lanes_df))\n",
    "    print(\"Intensity - Original points: \", len(pointcloud_df))\n",
    "    print(\"Intensity - Reduction to %:  \", len(lanes_df)/len(pointcloud_df))\n",
    "    \n",
    "    return lanes_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_trajectory_line(trajectory_file='./data/trajectory.fuse', min_x=0.0, min_y=0.0, min_z=0.0):\n",
    "    trajectory_df = read_fuse_to_df(trajectory_file)\n",
    "    trajectory_df, (min_x, min_y, min_z), (number, letter) = convert_fuse(trajectory_df, min_x, min_y, min_z)\n",
    "    trajectory_df[[\"Easting\", \"Northing\", \"Altitude\", \"Intensity\"]].to_csv(\"./data/trajectory.xyz\", index=False)\n",
    "    print(trajectory_df[[\"Easting\", \"Northing\", \"Altitude\", \"Intensity\"]])\n",
    "\n",
    "    line = \"LINESTRING(\"\n",
    "    for index, row in trajectory_df.iterrows():\n",
    "        line = line + str(row[\"Easting\"]) + \" \" + str(row[\"Northing\"]) + \" \" + str(row[\"Altitude\"]) + \", \"\n",
    "    line = line[:-2] + \")\"\n",
    "    trajectory_line = wkt.loads(line)\n",
    "    \n",
    "    return trajectory_line, (min_x, min_y, min_z), (number, letter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "def filter_by_trajectory_line(pointcloud_df, trajectory_line):\n",
    "    rows = []\n",
    "\n",
    "    for index, row in pointcloud_df.iterrows():\n",
    "        distance = trajectory_line.distance(row[\"Shape\"])\n",
    "        if distance <= 20:\n",
    "            rows.append(row)\n",
    "\n",
    "    filtered_lanes_df = pd.DataFrame(rows, columns=['Easting', 'Northing', 'Altitude', 'Intensity', 'Shape'])\n",
    "    \n",
    "    print(\"\\nTRAJECTORY FILTER:\")\n",
    "    print(\"============\")\n",
    "    print(\"Intensity - Filtered points: \", len(filtered_lanes_df))\n",
    "    print(\"Intensity - Original points: \", len(pointcloud_df))\n",
    "    print(\"Intensity - Reduction to %:  \", len(filtered_lanes_df)/len(pointcloud_df))\n",
    "    return filtered_lanes_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      Easting    Northing  Altitude  Intensity\n",
      "0   80.666066   33.179204    5.7431        255\n",
      "1   86.068856   44.074984    5.7281        255\n",
      "2   89.930128   51.890935    5.7252        255\n",
      "3   93.998028   60.187973    5.7307        255\n",
      "4   99.339125   71.096650    5.7376        255\n",
      "5  105.254758   83.187282    5.7454        255\n",
      "6  110.617644   94.092080    5.7775        255\n",
      "7  114.134203  101.207774    5.7856        255\n",
      "8  119.562210  112.114238    5.8083        255\n",
      "MEAN FILTER:\n",
      "============\n",
      "Intensity - Mean value:       8.246153096095984\n",
      "Intensity - Std value:        10.862899760737132\n",
      "Intensity - Lower bound:      19.109052856833117\n",
      "Intensity - Upper bound:      84.2864514212559\n",
      "Intensity - Filtered points:  32191\n",
      "Intensity - Original points:  430736\n",
      "Intensity - Reduction to %:   0.07473487240444263\n",
      "\n",
      "TRAJECTORY FILTER:\n",
      "============\n",
      "Intensity - Filtered points:  18496\n",
      "Intensity - Original points:  32191\n",
      "Intensity - Reduction to %:   0.5745705321363114\n"
     ]
    }
   ],
   "source": [
    "trajectory_line, (min_x, min_y, min_z), (number, letter) = get_trajectory_line('./data/trajectory.fuse',min_x, min_y, min_z)\n",
    "\n",
    "lanes_df = xyzi_df.copy()\n",
    "lanes_df = filter_by_mean_value(lanes_df, 1, 7)\n",
    "lanes_df[['Easting', 'Northing', 'Altitude', 'Intensity']].to_csv(\"./data/filter_mean.xyz\", index=False)\n",
    "lanes_df['Shape'] = lanes_df.apply(convert_to_shape, axis=1)\n",
    "lanes_df = filter_by_trajectory_line(lanes_df, trajectory_line)\n",
    "lanes_df[['Easting', 'Northing', 'Altitude', 'Intensity']].to_csv(\"./data/filter_trajectory.xyz\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Lane Marking Detection\n",
    "\n",
    "##### Idea\n",
    "After filtering the pointcloud with our trajectory filter and the mean value filter we can already see the lane markings. Unfortunately, we can also see the lane markings of the road into the other direction. However, you can seem that there are less points for these lane markings. Obviosly the density of points in a small area can be used to identify larger structures such as lane markings. Therefore, we wanted to group points that might belong together. We did this by using a Scikit Learn clustering algorithm. The DBSCAN algorithm allows us to specify a number for the points that are required to form a cluster and the distance between clusters. In our case this means that we require 40 points to build a cluster and the distance to be counted towards a cluster is 0.05. The result of the clustering algorithm can be seen in the following figure. Another side effect of the clustering is that we can get rid of \n",
    "4073 more points so that we now only have 14423 points that are most likely to be part of a lane marking.\n",
    "\n",
    "After clustering the points we can already see the lines. To gather actual line segments from the data we took the two points within each cluster that are furthest apart from each other. These two points serve as the starting and end point for a line segment. We use efficient numpy and scipy methods to calculate the distances in a quick manner. The result of the calculation can be visualized using matplotlib. The rendered example is illustrated by figure 2 in the results section below. The last step is then fusing some of the line segments that are quite close together. This allows us to increase the line length of some segments. This step is visualized in the third figure in the results section below.\n",
    "\n",
    "##### Problems\n",
    "An overall problem at this step ws aligning the line segments to the road trajectory. Since we do not know if the current road segment is very curvy or just straight we cannot assume that the road is straight all the time. However, our current algorithm assumes that the LIDAR pointcloud only represents a short segment 20 - 30 meters so that we might be able to solve this problem. Another problem is that some of the line segments are not straight or completely parallel to the trajectory. We plan to correct these issues by aligning all segments to the trajectory which should also allow us to handle curvatures.\n",
    "\n",
    "##### Results\n",
    "![Clustering LIDAR Pointcloud](Clustering.png \"Clustering LIDAR Pointcloud\")\n",
    "\n",
    "![Lines LIDAR Pointcloud](Lines.png \"Lines LIDAR Pointcloud\")\n",
    "\n",
    "![Fused LIDAR Pointcloud Lines](Fused.png \"Fused LIDAR Pointcloud Lines\")\n",
    "\n",
    "##### Code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Easting</th>\n",
       "      <th>Northing</th>\n",
       "      <th>Altitude</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>369</th>\n",
       "      <td>92.874609</td>\n",
       "      <td>37.312992</td>\n",
       "      <td>3.0119</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399</th>\n",
       "      <td>92.455209</td>\n",
       "      <td>36.957684</td>\n",
       "      <td>3.0039</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>496</th>\n",
       "      <td>92.065702</td>\n",
       "      <td>36.098402</td>\n",
       "      <td>3.0221</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>500</th>\n",
       "      <td>92.023889</td>\n",
       "      <td>36.063986</td>\n",
       "      <td>3.0189</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>916</th>\n",
       "      <td>91.028105</td>\n",
       "      <td>33.621722</td>\n",
       "      <td>3.0236</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Easting   Northing  Altitude  Group\n",
       "369  92.874609  37.312992    3.0119      0\n",
       "399  92.455209  36.957684    3.0039      0\n",
       "496  92.065702  36.098402    3.0221      0\n",
       "500  92.023889  36.063986    3.0189      0\n",
       "916  91.028105  33.621722    3.0236      1"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = lanes_df[[\"Easting\", \"Northing\", \"Altitude\"]].values\n",
    "X = StandardScaler().fit_transform(X)\n",
    "\n",
    "db = DBSCAN(eps=0.06, min_samples=30).fit(X)\n",
    "labels = db.labels_\n",
    "n_clusters_ = len(set(labels)) - (1 if -1 in labels else 0)\n",
    "n_noise_ = list(labels).count(-1)\n",
    "\n",
    "lanes_df[\"Group\"] = labels\n",
    "cluster_df = lanes_df[[\"Easting\", \"Northing\", \"Altitude\", \"Group\"]]\n",
    "cluster_df = cluster_df[cluster_df[\"Group\"]>=0]\n",
    "cluster_df.to_csv(\"./data/clustering.xyz\",index=False)\n",
    "cluster_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Final list of labels:        [-1 -1 -1 ...  4  5  4]\n",
      "Number of Clusters:          24\n",
      "Number of Noise points:      3435\n",
      "Number of clustered points:  15061\n"
     ]
    }
   ],
   "source": [
    "print(\"Final list of labels:       \", labels)\n",
    "print(\"Number of Clusters:         \", n_clusters_)\n",
    "print(\"Number of Noise points:     \", n_noise_)\n",
    "print(\"Number of clustered points: \", len(cluster_df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "lines = []\n",
    "\n",
    "for cluster in range(n_clusters_):\n",
    "    sub_cluster_df = cluster_df[cluster_df[\"Group\"] == cluster]\n",
    "    points = sub_cluster_df[[\"Easting\",\"Northing\", \"Altitude\"]].values\n",
    "    distances = squareform(pdist(points))\n",
    "    for i in range(0,15):\n",
    "        max_index = np.argmax(distances)\n",
    "        i1, i2 = np.unravel_index(max_index, distances.shape)\n",
    "        distances[i1,i2] = 0.0\n",
    "    max_dist = np.max(distances)\n",
    "    max_index = np.argmax(distances)\n",
    "    i1, i2 = np.unravel_index(max_index, distances.shape)\n",
    "    p1 = sub_cluster_df.iloc[i1]\n",
    "    p2 = sub_cluster_df.iloc[i2]\n",
    "    lines.append(([p1[\"Easting\"], p2[\"Easting\"]],[p1[\"Northing\"], p2[\"Northing\"]], [p1[\"Altitude\"], p2[\"Altitude\"]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of prototype lane markings:  24\n"
     ]
    }
   ],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.xlim(70, 130), plt.ylim(20,120)\n",
    "\n",
    "for l in lines:\n",
    "    plt.plot(l[0], l[1], l[2])\n",
    "    \n",
    "plt.show()\n",
    "\n",
    "print(\"Number of prototype lane markings: \", len(lines))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_line(coords):\n",
    "    return wkt.loads(\"LINESTRING(\" + str(coords[0][0]) + \" \" +str(coords[1][0]) + \", \" + str(coords[0][1]) + \" \" + str(coords[1][1]) +\")\")\n",
    "    \n",
    "distance_matrix = []\n",
    "\n",
    "for l1 in range(len(lines)):\n",
    "    prototype = []\n",
    "    distances = []\n",
    "    line1 = create_line(lines[l1])\n",
    "    for l2 in range(len(lines)):\n",
    "        line2 = create_line(lines[l2])\n",
    "        d = line1.distance(line2)\n",
    "        if d == 0:\n",
    "            d = 100\n",
    "        distances.append(d)\n",
    "        \n",
    "    distance_matrix.append(distances)\n",
    "\n",
    "distances = np.array(distance_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "fusions = set()\n",
    "\n",
    "for row in range(distances.shape[0]):\n",
    "    for column in range(distances.shape[1]):\n",
    "        value = distances[row,column]\n",
    "        if value < 1.0:\n",
    "            min_index = min(row, column)\n",
    "            max_index = max(row, column)\n",
    "            fusions.add((min_index, max_index))\n",
    "            \n",
    "fusions = list(fusions)\n",
    "fusion_clusters = []\n",
    "finished = []\n",
    "fusions.sort()\n",
    "\n",
    "for i in range(len(fusions)):\n",
    "    new_cluster = list(fusions[i])\n",
    "    if new_cluster[0] not in finished and new_cluster[1] not in finished:\n",
    "        for j in range(i+1, len(fusions)):\n",
    "            f = list(fusions[j])\n",
    "            if (f[0] in new_cluster or f[1] in new_cluster) and f[0] not in finished:\n",
    "                new_cluster.extend(f)\n",
    "        fusion_clusters.append(list(set(new_cluster)))\n",
    "        finished.extend(list(set(new_cluster)))\n",
    "        \n",
    "fused_lines = [item for sublist in fusion_clusters for item in sublist]\n",
    "for i in range(distances.shape[0]):\n",
    "    if i not in fused_lines:\n",
    "        fusion_clusters.append([i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_lane_markings = []\n",
    "\n",
    "for cluster in fusion_clusters:\n",
    "    if len(cluster) == 1:\n",
    "        final_lane_markings.append(lines[cluster[0]])\n",
    "    else:\n",
    "        points = []\n",
    "        for segment in cluster:\n",
    "            l = lines[segment]\n",
    "            points.append([l[0][0], l[1][0], l[2][0]])\n",
    "            points.append([l[0][1], l[1][1], l[2][1]])\n",
    "        points = np.array(points)\n",
    "        distances = squareform(pdist(points))\n",
    "        max_dist = np.max(distances)\n",
    "        max_index = np.argmax(distances)\n",
    "        i1, i2 = np.unravel_index(max_index, distances.shape)\n",
    "        \n",
    "        p1 = points[i1]\n",
    "        p2 = points[i2]\n",
    "\n",
    "        final_lane_markings.append(([p1[0], p2[0]],[p1[1], p2[1]], [p1[2], p2[2]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of final lane markings:  22\n"
     ]
    }
   ],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.xlim(70, 130), plt.ylim(20,120)\n",
    "\n",
    "for l in final_lane_markings:\n",
    "    plt.plot(l[0], l[1], l[2])\n",
    "    \n",
    "plt.show()\n",
    "\n",
    "print(\"Number of final lane markings: \", len(final_lane_markings))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Correction and Output\n",
    "\n",
    "##### Idea\n",
    "\n",
    "For the output the only thing that we needed to do is convert the UTM coordinates back to the latitude and longitude format. In the beginning we also shifted all the coordinates closer to the center of the coordinate system which means that we first have to shift all the UTM coordinates back to their original/true location. By using the python utm package we can easily transform the coordinates back into the original format. We then save the final output in a csv file. Each row contains a start and an end location for a line segment. That means that we will have a total of 22 rows + column headers.\n",
    "\n",
    "##### Results\n",
    "\n",
    "![Output CSV Final Lane Detections](Output.png \"Output CSV Final Lane Detections\")\n",
    "\n",
    "##### Code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output_rows = []\n",
    "\n",
    "for line in final_lane_markings:\n",
    "    start_x = line[0][0] + min_x\n",
    "    start_y = line[1][0] + min_y\n",
    "    start_z = line[2][0] + min_z\n",
    "    end_x = line[0][1] + min_x\n",
    "    end_y = line[1][1] + min_y\n",
    "    end_z = line[2][1] + min_z\n",
    "    (start_latitude, start_longitude) = utm.to_latlon(start_x,start_y,number, letter)\n",
    "    (end_latitude, end_longitude) = utm.to_latlon(end_x,end_y,number, letter)\n",
    "    output_rows.append([start_latitude, start_longitude, start_z, end_latitude, end_longitude, end_z])\n",
    "    \n",
    "output = pd.DataFrame(output_rows, columns=[\"Start_Latitude\", \"Start_Longitude\", \"Start_Z\",\"End_Latitude\", \"End_Longitude\", \"End_Z\"])\n",
    "output.to_csv(\"./data/output.csv\", header=True, index=False)"
   ]
  },
  {
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}