rectify_CACO_labels.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Rectify all images on S3 bucket and write to public read bucket"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\csherwood\\AppData\\Local\\Continuum\\miniconda3\\envs\\CRS\\lib\\site-packages\\skimage\\io\\manage_plugins.py:23: UserWarning: Your installed pillow version is < 7.1.0. Several security issues (CVE-2020-11538, CVE-2020-10379, CVE-2020-10994, CVE-2020-10177) have been fixed in pillow 7.1.0 or higher. We recommend to upgrade this library.\n",
      "  from .collection import imread_collection_wrapper\n"
     ]
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import imageio\n",
    "import fsspec\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import datetime\n",
    "from dateutil import tz\n",
    "import pandas as pd\n",
    "import glob, os\n",
    "\n",
    "from coastcam_funcs import *\n",
    "from calibration_crs import *\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Read the inventory of products"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['C:\\\\crs\\\\src\\\\dash_doodler\\\\results\\\\results2021-02-20-15-46/1577188801.c1.timex_label_greyscale2021-02-20-15-49_crs.png', 'C:\\\\crs\\\\src\\\\dash_doodler\\\\results\\\\results2021-02-20-15-46/1577188801.c2.timex_label_greyscale2021-02-20-15-53_crs.png']\n"
     ]
    }
   ],
   "source": [
    "rdir = 'C:\\\\crs\\\\src\\\\dash_doodler\\\\results\\\\results2021-02-20-15-46'\n",
    "c1list = []\n",
    "c2list = []\n",
    "\n",
    "for c1path in glob.glob(rdir+'/'+'*c1.timex_label_greyscale*'):\n",
    "    c1file = os.path.basename(c1path)\n",
    "\n",
    "    # check to see there is a corresponding c2 file\n",
    "    c2path = (glob.glob(rdir+'/'+c1file[0:10]+'.c2.timex_label_greyscale*'))[0]\n",
    "    if c2path:     \n",
    "        c1list.append(c1file)\n",
    "        c2file = os.path.basename(c2path)\n",
    "        c2list.append(c2file)\n",
    "\n",
    "label_paths = [ rdir+'/'+c1list[0],rdir+'/'+c2list[0]]\n",
    "print(label_paths)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# this is from ../src/dash_doodler/classes.txt\n",
    "label_names = ['sky','water','surf','veg','sand']\n",
    "label_clrs = ['lightblue','blue','gray','green','orange']\n",
    "    \n",
    "def sort_labels(labels):\n",
    "    label_nums = np.unique(labels)\n",
    "    label_size = np.zeros(len(label_nums))\n",
    "    for i,label_num in enumerate(label_nums):\n",
    "        label_size[i]= int(len(labels[labels==label_num]))\n",
    "    \n",
    "    sorted_labels = label_nums[np.argsort(-label_size)]\n",
    "    sorted_size = label_size[np.argsort(-label_size)]\n",
    "    return sorted_labels, sorted_size\n",
    "\n",
    "def label_inventory(labels, label_names):\n",
    "    sorted_labels, sorted_size = sort_labels(labels)\n",
    "    for i in sorted_labels:\n",
    "        print('{:}, {:d}, {:.3f}'.format(label_names[i-1],int(sorted_size[i-1]), sorted_size[i-1]/(np.sum(sorted_size))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Camera 1:\n",
      "veg, 628463, 0.125\n",
      "sand, 232420, 0.046\n",
      "water, 969784, 0.193\n",
      "sky, 2432612, 0.485\n",
      "surf, 750225, 0.150\n",
      "Camera 2:\n",
      "veg, 296098, 0.059\n",
      "water, 907184, 0.181\n",
      "sand, 185363, 0.037\n",
      "sky, 2903696, 0.579\n",
      "surf, 721163, 0.144\n"
     ]
    }
   ],
   "source": [
    "print('Camera 1:')\n",
    "lbl = imageio.imread(label_paths[0])\n",
    "label_inventory(lbl,label_names)\n",
    "print('Camera 2:')\n",
    "lbl = imageio.imread(label_paths[1])\n",
    "label_inventory(lbl,label_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make a time series of the fraction for each label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Length of image list:  17\n",
      "Length of label list:  5\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x21bd6601ac8>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print('Length of image list: ',len(c1list))\n",
    "print('Length of label list: ',len(label_names))\n",
    "\n",
    "# Make an array w/ number of labels in each category\n",
    "labc = np.zeros((len(c1list),len(label_names)))\n",
    "for i, fn1 in enumerate(c1list):\n",
    "    lbl = imageio.imread( rdir+'/'+fn1)\n",
    "    sl, ss = sort_labels(lbl)\n",
    "    for j in range(len(sl)):\n",
    "        labc[i,sl[j]-1] = ss[j]\n",
    "        \n",
    "# normalize by number of labels per image\n",
    "nc = lbl.size\n",
    "labc = labc / nc\n",
    "        \n",
    "sky = labc[:,0]\n",
    "water = labc[:,1]\n",
    "surf = labc[:,2]\n",
    "veg = labc[:,3]\n",
    "sand = labc[:,4]\n",
    "fig,ax = plt.subplots(figsize=(7,5))\n",
    "plt.plot(veg,label='veg',linewidth=2,c=label_clrs[3])\n",
    "plt.plot(sand,label='sand',linewidth=2,c=label_clrs[4])\n",
    "plt.plot(water,label='water',linewidth=2,c=label_clrs[1])\n",
    "plt.plot(sky,label='sky',linewidth=2,c=label_clrs[0])\n",
    "plt.plot(surf,label='surf',linewidth=2,c=label_clrs[2])\n",
    "plt.ylabel('Fraction of oblique image')\n",
    "plt.xlabel('Image')\n",
    "plt.title('Camera 1')\n",
    "plt.legend()\n",
    "\n",
    "# Make an array w/ number of labels in each category\n",
    "labc = np.zeros((len(c2list),len(label_names)))\n",
    "for i, fn1 in enumerate(c2list):\n",
    "    lbl = imageio.imread( rdir+'/'+fn1)\n",
    "    sl, ss = sort_labels(lbl)\n",
    "    for j in range(len(sl)):\n",
    "        labc[i,sl[j]-1] = ss[j]\n",
    "        \n",
    "# normalize by number of labels per image\n",
    "nc = lbl.size\n",
    "labc = labc / nc\n",
    "        \n",
    "sky = labc[:,0]\n",
    "water = labc[:,1]\n",
    "surf = labc[:,2]\n",
    "veg = labc[:,3]\n",
    "sand = labc[:,4]\n",
    "fig,ax = plt.subplots(figsize=(7,5))\n",
    "plt.plot(veg,label='veg',linewidth=2,c=label_clrs[3])\n",
    "plt.plot(sand,label='sand',linewidth=2,c=label_clrs[4])\n",
    "plt.plot(water,label='water',linewidth=2,c=label_clrs[1])\n",
    "plt.plot(sky,label='sky',linewidth=2,c=label_clrs[0])\n",
    "plt.plot(surf,label='surf',linewidth=2,c=label_clrs[2])\n",
    "plt.ylabel('Fraction of oblique image')\n",
    "plt.xlabel('Image')\n",
    "plt.title('Camera 2')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Set up the rectification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extrinsics are in world coordinates\n",
      "{'x': 410843.97, 'y': 4655942.49, 'z': 27.3, 'a': -0.271, 't': 1.304, 'r': 0.007}\n",
      "52.49000000022352\n",
      "{'x': 410935.0, 'y': 4655890.0, 'angd': 55.0}\n",
      "{'x': 410843.97, 'y': 4655942.49, 'z': 27.3, 'a': -0.271, 't': 1.304, 'r': 0.007}\n",
      "{'x': -9.21537219613947, 'y': 104.6744377358444, 'z': 27.3, 'a': 0.6889310885968812, 't': 1.304, 'r': 0.007}\n"
     ]
    }
   ],
   "source": [
    "# List of files...three for each camera. Calibration parameters are in .json format\n",
    "# These are the USGS image filename format\n",
    "extrinsic_cal_files = ['CACO01_C1_EOBest.json','CACO01_C2_EOBest.json']\n",
    "intrinsic_cal_files = ['CACO01_C1_IOBest.json','CACO01_C2_IOBest.json']\n",
    "\n",
    "# Dict providing the metadata that the Axiom code infers from the USACE filename format\n",
    "metadata= {'name': 'CACO-01', 'serial_number': 1, 'camera_number': 'C1', 'calibration_date': '2019-12-12', 'coordinate_system': 'geo'}\n",
    "# dict providing origin and orientation of the local grid\n",
    "local_origin = {'x': 410935.,'y':4655890., 'angd': 55.}\n",
    "\n",
    "# read cal files and make lists of cal dicts\n",
    "extrinsics_list = []\n",
    "for f in extrinsic_cal_files:\n",
    "    extrinsics_list.append( json2dict(f) )\n",
    "intrinsics_list = []\n",
    "for f in intrinsic_cal_files:\n",
    "    intrinsics_list.append( json2dict(f) )\n",
    "    \n",
    "# check test for coordinate system\n",
    "if metadata['coordinate_system'].lower() == 'xyz':\n",
    "    print('Extrinsics are local coordinates')\n",
    "elif metadata['coordinate_system'].lower() == 'geo':\n",
    "    print('Extrinsics are in world coordinates')\n",
    "else:\n",
    "    print('Invalid value of coordinate_system: ',metadata['coordinate_system'])\n",
    "    \n",
    "print(extrinsics_list[0])\n",
    "print(extrinsics_list[0]['y']-local_origin['y'])\n",
    "\n",
    "calibration = CameraCalibration(metadata,intrinsics_list[0],extrinsics_list[0],local_origin)\n",
    "print(calibration.local_origin)\n",
    "print(calibration.world_extrinsics)\n",
    "print(calibration.local_extrinsics)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Functions for rectifying labels\n",
    "* TODO - Put these in an external .py file\n",
    "* `TargetGrid` is unchanged...remove from here if `rectify_crs.py` is included.\n",
    "* The `rectfier` class has been split into standalone functions, renamed, and repurposed to treat 2D arrays, rather than x,y,rgb arrays."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.interpolate import RegularGridInterpolator\n",
    "from scipy.ndimage.morphology import distance_transform_edt\n",
    "\n",
    "class TargetGrid(object):\n",
    "    \"\"\"Grid generated to georectify image.\n",
    "\n",
    "    CRS modified to make endpoints inclusive\n",
    "    Notes:\n",
    "        - Used to maps points in world coordinates to pixels\n",
    "        - The limits should be specified in local coordinates using the same\n",
    "          coordinates and units as camera calibrations.\n",
    "\n",
    "    Args:\n",
    "        xlims (ndarray) - min and max (inclusive) in the x-direction (e.g. [-50, 650])\n",
    "        ylims (ndarray) - min and max (inclusive) in the y-direction (e.g. [0, 2501])\n",
    "        dx (float) - resolution of grid in x direction (same units as camera calibration)\n",
    "        dy (float) - resolution of grid in y direction (same units as camera calibration)\n",
    "        z (float) - static value to estimate elevation at everypoint in the x, y grid\n",
    "\n",
    "    Attributes:\n",
    "        X (np.ndarray): Local grid coordinates in x-direction.\n",
    "        Y (np.ndarray): Local grid coordinates in y-direction.\n",
    "        Z (np.ndarray): Local grid coordinates in z-direction.\n",
    "        xyz (np.ndarray): The grid where pixels are compiled from images for rectification.\n",
    "    \"\"\"\n",
    "    def __init__(self, xlims, ylims, dx=1, dy=1, z=-0.91):\n",
    "        x = np.arange(xlims[0], xlims[1]+dx, dx)\n",
    "        y = np.arange(ylims[0], ylims[1]+dx, dy)\n",
    "        self.X, self.Y = np.meshgrid(x, y)\n",
    "        self.Z = np.zeros_like(self.X) + z\n",
    "        self.xyz = self._xyz_grid()\n",
    "\n",
    "    def _xyz_grid(self):\n",
    "        x = self.X.copy().T.flatten()\n",
    "        y = self.Y.copy().T.flatten()\n",
    "        z = self.Z.copy().T.flatten()\n",
    "        return np.vstack((x, y, z)).T\n",
    "    \n",
    "def find_distort_UV(target_grid, calibration):\n",
    "    # get UV for pinhole camera\n",
    "    xyz = np.vstack((\n",
    "        target_grid.xyz.T,\n",
    "        np.ones((len(target_grid.xyz),))\n",
    "    ))\n",
    "    UV = np.matmul(calibration.P, xyz)\n",
    "\n",
    "    # make homogenous\n",
    "    div = np.tile(UV[2, :], (3, 1))\n",
    "    UV = UV / div\n",
    "\n",
    "    # get and rename\n",
    "    NU = calibration.lcp['NU']\n",
    "    NV = calibration.lcp['NV']\n",
    "    c0U = calibration.lcp['c0U']\n",
    "    c0V = calibration.lcp['c0V']\n",
    "    fx = calibration.lcp['fx']\n",
    "    fy = calibration.lcp['fy']\n",
    "    d1 = calibration.lcp['d1']\n",
    "    d2 = calibration.lcp['d2']\n",
    "    d3 = calibration.lcp['d3']\n",
    "    t1 = calibration.lcp['t1']\n",
    "    t2 = calibration.lcp['t2']\n",
    "    u = UV[0, :]\n",
    "    v = UV[1, :]\n",
    "\n",
    "    # normalize distances\n",
    "    x = (u - c0U) / fx\n",
    "    y = (v - c0V) / fy\n",
    "    # radial distortion\n",
    "    r2 = x*x + y*y\n",
    "    fr = 1. + d1*r2 + d2*r2*r2 + d3*r2*r2*r2\n",
    "    # tangential distorion\n",
    "    dx=2.*t1*x*y + t2*(r2+2.*x*x)\n",
    "    dy=t1*(r2+2.*y*y) + 2.*t2*x*y\n",
    "    # apply correction, answer in chip pixel units\n",
    "    xd = x*fr + dx\n",
    "    yd = y*fr + dy\n",
    "    Ud = xd*fx+c0U\n",
    "    Vd = yd*fy+c0V\n",
    "\n",
    "    # Declare array for flagged values\n",
    "    flag = np.ones_like(Ud)\n",
    "\n",
    "    # find negative UV coordinates\n",
    "    flag[np.where( Ud<0.)]=0.\n",
    "    flag[np.where( Vd<0.)]=0.\n",
    "    # find UVd coordinates greater than image size\n",
    "    flag[np.where( Ud>=NU)]=0.\n",
    "    flag[np.where( Vd>=NV)]=0.\n",
    "\n",
    "    # Determine if Tangential Distortion is within Range\n",
    "    #  Find Maximum possible tangential distortion at corners\n",
    "    Um=np.array((0, 0, NU, NU))\n",
    "    Vm=np.array((0, NV, NV, 0))\n",
    "\n",
    "    # Normalization\n",
    "    xm = (Um-c0U)/fx\n",
    "    ym = (Vm-c0V)/fy\n",
    "    r2m = xm*xm + ym*ym\n",
    "\n",
    "    # Tangential Distortion\n",
    "    dxm=2.*t1*xm*ym + t2*(r2m+2.*xm*xm)\n",
    "    dym=t1*(r2m+2.*ym*ym) + 2.*t2*xm*ym\n",
    "\n",
    "    # Find Values Larger than those at corners\n",
    "    flag[np.where(np.abs(dy)>np.max(np.abs(dym)))]=0.\n",
    "    flag[np.where(np.abs(dx)>np.max(np.abs(dxm)))]=0.\n",
    "\n",
    "    DU = Ud.reshape(target_grid.X.shape, order='F')\n",
    "    DV = Vd.reshape(target_grid.Y.shape, order='F')\n",
    "\n",
    "    # find negative Zc values and add to flag\n",
    "    UV = np.matmul(calibration.P, xyz)\n",
    "    xyzC = np.matmul(calibration.R,np.matmul(calibration.IC,xyz))\n",
    "    flag[np.where(xyzC[2,:]<=0.)]=0.\n",
    "    flag = flag.reshape(target_grid.X.shape, order='F')\n",
    "\n",
    "    # apply the flag to zero-out non-valid points\n",
    "    return DU*flag, DV*flag, flag\n",
    "\n",
    "def get_labels(target_grid, DU, DV, labels):\n",
    "    \"\"\"Return label values for each xyz point from the image\n",
    "\n",
    "    Arguments:\n",
    "        DU (np.ndarray): Pixel location in camera orientation and coordinate system\n",
    "        DV (np.ndarray): Pixel location in camera orientation and coorindate system\n",
    "        image_labels (np.ndarray [nx,ny]) with integer label values at U,V points\n",
    "        interpolation will be nearest neighbor to preserve integer values\n",
    "\n",
    "    Returns:\n",
    "        K (np.ndarray): Integer image label for each point in the image\n",
    "    \"\"\"\n",
    "    K = np.zeros((\n",
    "        target_grid.X.shape[0],\n",
    "        target_grid.X.shape[1]\n",
    "    ))\n",
    "\n",
    "    rgi = RegularGridInterpolator(\n",
    "        (np.arange(0, labels.shape[0]),\n",
    "         np.arange(0, labels.shape[1])),\n",
    "        labels[:,:],\n",
    "        method='nearest',\n",
    "        bounds_error=False,\n",
    "        fill_value=0)\n",
    "    K[:, :] = rgi((DV,DU))\n",
    "\n",
    "    # mask out values out of range like Matlab\n",
    "    # avoid runtime nan comparison warning (DU, DV already have nans)\n",
    "    with np.errstate(invalid='ignore'):\n",
    "        mask_u = np.logical_or(\n",
    "            DU <= 1,\n",
    "            DU >= labels.shape[1]\n",
    "        )\n",
    "        mask_v = np.logical_or(\n",
    "            DV <= 1,\n",
    "            DV >= labels.shape[0]\n",
    "        )\n",
    "    mask = np.logical_or(\n",
    "        mask_u,\n",
    "        mask_v\n",
    "    )\n",
    "    K[mask] = np.nan\n",
    "    return K\n",
    "\n",
    "def rectify_labels(metadata, target_grid, label_files, intrinsic_cal_list, extrinsic_cal_list, local_origin, fs=None):\n",
    "    \"\"\"Georectify and blend labels from multiple cameras\n",
    "\n",
    "    Arguments:\n",
    "        metadata (dict):\n",
    "        target_grid: grid object\n",
    "        label_files (list): List of label files\n",
    "        intrinsic_cal_list (list): list of paths to internal calibrations (one for each camera)\n",
    "        extrinsic_cal_list (list): list of paths to external calibrations (one for each camera)\n",
    "        local_origin:\n",
    "        fs: (object): fsspec file spec object for folder on S3 bucket. If none, normal file system will be used.\n",
    "\n",
    "    Returns:\n",
    "        M (np.ndarray): Georectified images merged from supplied images.\n",
    "    \"\"\"\n",
    "    # array for final pixel values\n",
    "    #M = np.zeros_like(target_grid.X)\n",
    "    M = np.zeros((701,501,2))\n",
    "\n",
    "    for cur_idx, (label_file, intrinsic_cal, extrinsic_cal) in enumerate(zip(label_files, intrinsic_cal_list, extrinsic_cal_list)):\n",
    "        # load camera calibration file and find pixel locations\n",
    "        camera_calibration = CameraCalibration(metadata, intrinsic_cal, extrinsic_cal, local_origin)\n",
    "        U, V, flag = find_distort_UV(target_grid, camera_calibration)\n",
    "\n",
    "        # load image and apply weights to pixels\n",
    "        if fs:\n",
    "            # using fsspec for S3 files\n",
    "            with fs.open(label_file) as f:\n",
    "                image = imageio.imread(f)\n",
    "        else:\n",
    "            # regular file system\n",
    "            image = imageio.imread(label_file)\n",
    "\n",
    "        M[:,:,cur_idx] = get_labels(target_grid, U, V, image)\n",
    "    \n",
    "    return M.astype(np.uint8)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "xmin = 0.\n",
    "xmax = 500.\n",
    "ymin = 0.\n",
    "ymax = 700.\n",
    "dx = 1.\n",
    "dy = 1.\n",
    "z =  0.\n",
    "\n",
    "rectifier_grid = TargetGrid(\n",
    "    [xmin, xmax],\n",
    "    [ymin, ymax],\n",
    "    dx,\n",
    "    dy,\n",
    "    z\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5\n"
     ]
    }
   ],
   "source": [
    "M = rectify_labels(metadata, rectifier_grid, label_paths, intrinsics_list, \\\n",
    "                                                extrinsics_list, local_origin )        \n",
    "Mm = np.maximum(M[:,:,0],M[:,:,1])\n",
    "\n",
    "print(np.max(Mm))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.colors import ListedColormap\n",
    "label_cmap = ListedColormap(['#696969','#00FFFF', '#0000FF', '#F5FFFA', '#008000', '#FFA500'])\n",
    "cb_label_names = ['none','sky','water','surf','veg','sand']\n",
    "\n",
    "fig = plt.figure(figsize=(7,5))\n",
    "# fig,ax = plt.subplots(1,1,aspect='equal')\n",
    "plt.pcolor(Mm,cmap=label_cmap,)\n",
    "plt.gca().set_aspect('equal')\n",
    "cb=plt.colorbar()\n",
    "cb.set_ticks(np.linspace(.5, 4.5, 6))\n",
    "cb.set_ticklabels(cb_label_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1152 with 20 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "i = 0\n",
    "fig,ax=plt.subplots(5,4,figsize=(20,16))\n",
    "ax = ax.ravel()\n",
    "for fn1, fn2 in (zip(c1list,c2list)):\n",
    "    label_paths = [ rdir+'/'+fn1, rdir+'/'+fn2 ]\n",
    "    M = rectify_labels(metadata, rectifier_grid, label_paths, intrinsics_list, \\\n",
    "                                                extrinsics_list, local_origin )        \n",
    "    Mm = np.maximum(M[:,:,0],M[:,:,1])\n",
    "    Mm[Mm==1]=3\n",
    "    #ax[i].subplot(5,4,i+1)\n",
    "    ax[i].pcolor(Mm,cmap=label_cmap)\n",
    "    ax[i].set_xticks([])\n",
    "    ax[i].set_yticks([])\n",
    "    ax[i].set_aspect('equal')\n",
    "    i=i+1"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}