methods.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import cv2\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "import matplotlib.image as mpimg\n",
    "import glob\n",
    "import time\n",
    "\n",
    "import scipy.misc\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.externals import joblib\n",
    "from sklearn.model_selection import train_test_split\n",
    "from skimage.feature import hog"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot3d(pixels, colors_rgb,\n",
    "        axis_labels=list(\"RGB\"), axis_limits=[(0, 255), (0, 255), (0, 255)]):\n",
    "    \"\"\"Plot pixels in 3D.\"\"\"\n",
    "\n",
    "    # Create figure and 3D axes\n",
    "    fig = plt.figure(figsize=(8, 8))\n",
    "    ax = Axes3D(fig)\n",
    "\n",
    "    # Set axis limits\n",
    "    ax.set_xlim(*axis_limits[0])\n",
    "    ax.set_ylim(*axis_limits[1])\n",
    "    ax.set_zlim(*axis_limits[2])\n",
    "\n",
    "    # Set axis labels and sizes\n",
    "    ax.tick_params(axis='both', which='major', labelsize=14, pad=8)\n",
    "    ax.set_xlabel(axis_labels[0], fontsize=16, labelpad=16)\n",
    "    ax.set_ylabel(axis_labels[1], fontsize=16, labelpad=16)\n",
    "    ax.set_zlabel(axis_labels[2], fontsize=16, labelpad=16)\n",
    "\n",
    "    # Plot pixel values with colors given in colors_rgb\n",
    "    ax.scatter(\n",
    "        pixels[:, :, 0].ravel(),\n",
    "        pixels[:, :, 1].ravel(),\n",
    "        pixels[:, :, 2].ravel(),\n",
    "        c=colors_rgb.reshape((-1, 3)), edgecolors='none')\n",
    "\n",
    "    return ax  # return Axes3D object for further manipulation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Define a function to compute binned color features  \n",
    "def bin_spatial(img, size=(32, 32)):\n",
    "    # Use cv2.resize().ravel() to create the feature vector\n",
    "    features = cv2.resize(img, size).ravel() \n",
    "    # Return the feature vector\n",
    "    return features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Define a function to compute color histogram features  \n",
    "def color_hist(img, nbins=32, bins_range=(0, 256)):\n",
    "    # Compute the histogram of the color channels separately\n",
    "    channel1_hist = np.histogram(img[:,:,0], bins=nbins, range=bins_range)\n",
    "    channel2_hist = np.histogram(img[:,:,1], bins=nbins, range=bins_range)\n",
    "    channel3_hist = np.histogram(img[:,:,2], bins=nbins, range=bins_range)\n",
    "    # Concatenate the histograms into a single feature vector\n",
    "    hist_features = np.concatenate((channel1_hist[0], channel2_hist[0], channel3_hist[0]))\n",
    "    # Return the individual histograms, bin_centers and feature vector\n",
    "    return hist_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Define a function to return HOG features and visualization\n",
    "def get_hog_features(img, orient=9, pix_per_cell=8, cell_per_block=2,\n",
    "                        vis=False, feature_vec=True):\n",
    "    # Call with two outputs if vis==True\n",
    "    if vis == True:\n",
    "        features, hog_image = hog(img, orientations=orient, pixels_per_cell=(pix_per_cell, pix_per_cell),\n",
    "                                  cells_per_block=(cell_per_block, cell_per_block), transform_sqrt=True, \n",
    "                                  visualise=vis, feature_vector=feature_vec)\n",
    "        return features, hog_image\n",
    "    # Otherwise call with one output\n",
    "    else:      \n",
    "        features = hog(img, orientations=orient, pixels_per_cell=(pix_per_cell, pix_per_cell),\n",
    "                       cells_per_block=(cell_per_block, cell_per_block), transform_sqrt=True, \n",
    "                       visualise=vis, feature_vector=feature_vec)\n",
    "        return features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def extract_features_img(img,cspace='RGB', spatial_size=(32, 32),\n",
    "                        hist_bins=32, hist_range=(0, 256), orient=9, \n",
    "                        pix_per_cell=8, cell_per_block=2, hog_channel=0,debug=0):\n",
    "# apply color conversion if other than 'RGB'\n",
    "        if cspace != 'RGB':\n",
    "            if cspace == 'HSV':\n",
    "                feature_image = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)\n",
    "            elif cspace == 'LUV':\n",
    "                feature_image = cv2.cvtColor(img, cv2.COLOR_BGR2LUV)\n",
    "            elif cspace == 'HLS':\n",
    "                feature_image = cv2.cvtColor(img, cv2.COLOR_BGR2HLS)\n",
    "            elif cspace == 'YUV':\n",
    "                feature_image = cv2.cvtColor(img, cv2.COLOR_BGR2YUV)\n",
    "            elif cspace == 'YCrCb':\n",
    "                feature_image = cv2.cvtColor(img, cv2.COLOR_RGB2YCrCb)\n",
    "        else: \n",
    "            image = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) # convert to rgb\n",
    "            feature_image = np.copy(image)\n",
    "    \n",
    "        t1 = time.time()\n",
    "        # Apply bin_spatial() to get spatial color features\n",
    "        spatial_features = bin_spatial(feature_image, size=spatial_size)\n",
    "        \n",
    "        t2 = time.time()\n",
    "        # Apply color_hist() also with a color space option now\n",
    "        hist_features = color_hist(feature_image, nbins=hist_bins, bins_range=hist_range)\n",
    "        \n",
    "        t3 = time.time()\n",
    "        # Call get_hog_features() with vis=False, feature_vec=True\n",
    "        hog_features = get_hog_features(feature_image[:,:,hog_channel], orient, \n",
    "                                        pix_per_cell, cell_per_block, vis=False, feature_vec=True)\n",
    "        t4 = time.time()\n",
    "        \n",
    "        if debug:   \n",
    "            print ('spatial_time = ',(t2-t1)*1000,' ms')\n",
    "            print ('histogram_time = ',(t3-t2)*1000,' ms')\n",
    "            print ('hog_time = ',(t4-t3)*1000,' ms')\n",
    "            hog_features_0, hog_img_0= get_hog_features(feature_image[:,:,0], orient, \n",
    "                                        pix_per_cell, cell_per_block, vis=True, feature_vec=True)\n",
    "            hog_features_1, hog_img_1= get_hog_features(feature_image[:,:,1], orient, \n",
    "                                        pix_per_cell, cell_per_block, vis=True, feature_vec=True)\n",
    "            hog_features_2, hog_img_2= get_hog_features(feature_image[:,:,2], orient, \n",
    "                                        pix_per_cell, cell_per_block, vis=True, feature_vec=True)\n",
    "#             cv2.imwrite('RGB_hog_0.png',hog_img_0)\n",
    "#             cv2.imwrite('RGB_hog_1.png',hog_img_1)\n",
    "#             cv2.imwrite('RGB_hog_2.png',hog_img_2)\n",
    "            \n",
    "            width = 1/1.5\n",
    "            fig, ax = plt.subplots( nrows=1, ncols=1 )  # create figure & 1 axis\n",
    "            fig.tight_layout();\n",
    "            ax.bar(range(len(spatial_features)), spatial_features, width, color=\"blue\")\n",
    "            filename = cspace+'_spatial_features'\n",
    "            fig.suptitle(filename, fontsize=12)\n",
    "            plt.show()\n",
    "            fig.savefig('output_images/'+filename+'.png')\n",
    "            ###########################################################################\n",
    "            fig, ax = plt.subplots( nrows=1, ncols=1 )  # create figure & 1 axis\n",
    "            fig.tight_layout();\n",
    "            ax.bar(range(len(hist_features)), hist_features, width, color=\"blue\")\n",
    "            filename = cspace+'_hist_features'\n",
    "            fig.suptitle(filename, fontsize=12)\n",
    "            plt.show()\n",
    "            fig.savefig('output_images/'+filename+'.png')\n",
    "            ###########################################################################\n",
    "            fig, ax = plt.subplots( nrows=1, ncols=1 )  # create figure & 1 axis\n",
    "            fig.tight_layout();\n",
    "            ax.bar(range(len(hog_features)), hog_features, width, color=\"blue\")\n",
    "            filename = cspace+'_hog_features'\n",
    "            fig.suptitle(filename, fontsize=12)\n",
    "            plt.show()\n",
    "            fig.savefig('output_images/'+filename+'.png')\n",
    "            ###########################################################################\n",
    "            fig, ax = plt.subplots( nrows=1, ncols=1 )  # create figure & 1 axis\n",
    "            fig.tight_layout();\n",
    "            ax.axis('off')\n",
    "            plt.imshow(hog_img_0,cmap='gray')\n",
    "            filename = cspace+'_hog_ch_0'\n",
    "            fig.suptitle(filename, fontsize=12)\n",
    "            plt.show()\n",
    "            fig.savefig('output_images/'+filename+'.png')\n",
    "            ###########################################################################\n",
    "            fig, ax = plt.subplots( nrows=1, ncols=1 )  # create figure & 1 axis\n",
    "            fig.tight_layout();\n",
    "            ax.axis('off')\n",
    "            plt.imshow(hog_img_1,cmap='gray')\n",
    "            filename = cspace+'_hog_ch_1'\n",
    "            fig.suptitle(filename, fontsize=12)\n",
    "            plt.show()\n",
    "            fig.savefig('output_images/'+filename+'.png')\n",
    "            ###########################################################################\n",
    "            fig, ax = plt.subplots( nrows=1, ncols=1 )  # create figure & 1 axis\n",
    "            fig.tight_layout();\n",
    "            ax.axis('off')\n",
    "            plt.imshow(hog_img_2,cmap='gray')\n",
    "            filename = cspace+'_hog_ch_2'\n",
    "            fig.suptitle(filename, fontsize=12)\n",
    "            plt.show()\n",
    "            fig.savefig('output_images/'+filename+'.png')\n",
    "\n",
    "        return np.concatenate((spatial_features, hist_features, hog_features))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Define a function to extract features from a list of images\n",
    "# Have this function call bin_spatial() and color_hist()\n",
    "# this function combines color, histogram and hog features extraction\n",
    "def extract_features(imgs, cspace='RGB', spatial_size=(32, 32),\n",
    "                        hist_bins=32, hist_range=(0, 256), orient=9, \n",
    "                        pix_per_cell=8, cell_per_block=2, hog_channel=0,debug=0):\n",
    "    # Create a list to append feature vectors to\n",
    "    features = []\n",
    "    # Iterate through the list of images\n",
    "    for file in imgs:\n",
    "        # Read in each one by one\n",
    "#         image = mpimg.imread(file)\n",
    "        image = cv2.imread(file) # reads a file into bgr values 0-255\n",
    "        img_features = extract_features_img(img,cspace='RGB',hog_channel=0,debug=debug)\n",
    "        # Append the new feature vector to the features list\n",
    "        features.append(img_features)\n",
    "        \n",
    "    # Return list of feature vectors\n",
    "    return features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Define a function that takes an image,\n",
    "# start and stop positions in both x and y, \n",
    "# window size (x and y dimensions),  \n",
    "# and overlap fraction (for both x and y)\n",
    "def slide_window(img, x_start_stop=[None, None], y_start_stop=[None, None], \n",
    "                    xy_window=(64, 64), xy_overlap=(0.5, 0.5),max_y=780):\n",
    "    height, width, channels = img.shape\n",
    "    # If x and/or y start/stop positions not defined, set to image size\n",
    "    if x_start_stop[0] == None:\n",
    "        x_start_stop[0] = 0\n",
    "    if x_start_stop[1] == None:\n",
    "        x_start_stop[1] = img.shape[1]\n",
    "    if y_start_stop[0] == None:\n",
    "        y_start_stop[0] = 0\n",
    "    if y_start_stop[1] == None:\n",
    "        y_start_stop[1] = img.shape[0]\n",
    "    # Compute the span of the region to be searched    \n",
    "    xspan = x_start_stop[1] - x_start_stop[0]\n",
    "    yspan = y_start_stop[1] - y_start_stop[0]\n",
    "    # Compute the number of pixels per step in x/y\n",
    "    nx_pix_per_step = np.int(xy_window[0]*(1 - xy_overlap[0]))\n",
    "    ny_pix_per_step = np.int(xy_window[1]*(1 - xy_overlap[1]))\n",
    "    # Compute the number of windows in x/y\n",
    "    nx_windows = np.int(xspan/nx_pix_per_step) - 1\n",
    "    ny_windows = np.int(yspan/ny_pix_per_step) - 1\n",
    "    # Initialize a list to append window positions to\n",
    "    window_list = []\n",
    "    # Loop through finding x and y window positions\n",
    "    # Note: you could vectorize this step, but in practice\n",
    "    # you'll be considering windows one by one with your\n",
    "    # classifier, so looping makes sense\n",
    "    for ys in range(ny_windows):\n",
    "        for xs in range(nx_windows):\n",
    "            # Calculate window position\n",
    "            startx = int(xs*nx_pix_per_step + x_start_stop[0])\n",
    "            endx = int(startx + xy_window[0])\n",
    "            starty = int(ys*ny_pix_per_step + y_start_stop[0])\n",
    "            endy = int(starty + xy_window[1])\n",
    "            # Append window position to list\n",
    "            if endy<height and endx < width and endy<max_y:\n",
    "                window_list.append(((startx, starty), (endx, endy)))\n",
    "    # Return the list of windows\n",
    "    return window_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "spatial_time =  0.013589859008789062  ms\n",
      "histogram_time =  0.5838871002197266  ms\n",
      "hog_time =  1.2407302856445312  ms\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAacAAAEdCAYAAACovqiLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGSFJREFUeJzt3XuQXOV95vHnQZIlrMtIMBLIgHVJhAC5Ashjlo1YYBev\njVSORVzGkRevFcAlO8FUspVUVibrgKviGxVnK9RmyeIyRtjCNo7jhWxIiIr4kmwisOTlIkVGGiSB\nxAzSDGAhhEs3fvtHvyOaoeei6e7p95z+fqq6+vTbp/u8vz6nz9Pn7Z4zjggBAJCTU1rdAQAABiOc\nAADZIZwAANkhnAAA2SGcAADZIZwAANkhnIA62d5q+8pRzhu2f3mEeU61/de2D9j+bkM6CRQM4YSm\nsr3b9i9sv2r7Bdv32J5WdX+X7f9j+2XbP7f9r7Y/b3tWuv83bR9Pj3/V9k7bv9XCeu6x/cfVbRGx\nJCJ+2MDFfFjSGZJOj4hr63ki27fZ/mZjugWMH8IJ4+HXImKapIskXSzpM5Jk+1cl/VDS/5V0XkTM\nlHS1pGOSLqx6/L9ExLT0HB+WdLvti8ex/+NtnqTtEXGs1R2xPbHVfUB7IpwwbiLiBUkPqxJSknS7\npK9HxBcjYl+a57mIuHWoI5GI+KmkbZLOH25ZtqfY/qbtF9MR2U9sn5Hu+6HtL9p+LA2dPWD7tKrH\nfjcd5R2w/WPbS1L7GknXSfqDdBT316l9t+33pulLbP9LWmav7f9h+22jfY1sf07SH0n6jbSMG1P7\nDba3pSPMh23Pq3rMn9neY/sV25tt/7vUfrWkW6qe64nB/U23Txxd2Z6fhh5vtP2cpH9I7Zfa/udU\n1xPVw5jp6Han7YO2d9m+brT1AkMhnDBubJ8tabmkbttTJf1bSd87yed4j6RzJW0aYdbVkjoknSPp\ndEmfkvSLqvs/LukGSe9Q5Ujtjqr7/lbSIklzJP1U0npJioi70vTt6Uju12os97ik/yKpU5X6rpL0\n26OtLyJulfQFSd9Jy/ia7WtUCZkPSZot6R8lfavqYT9RJfBPk3SfpO/anhIRfzfouaqPRkdyhSof\nAN5v+yxJfyPpj9Myfl/S92zPTuvxDknLI2K6pF+V9PhJLAeoiXDCePjftg9K2iNpv6RbJc1SZft7\nYWAm27enT+aHbP+3qsdfmtpflfSYpG9I2jHCMo+qEkq/HBHHI2JzRLxSdf83ImJLRByS9FlJH7E9\nQZIi4u6IOBgRhyXdJulC2x2jKTQtZ2NEHIuI3ZL+lyo7+np8UtIXI2JbGur7gqSLBo6eIuKbEfFi\nWuZXJE2WtLjOZd4WEYci4heSPibpoYh4KCJej4gNqnw4WJHmfV3Su2yfGhG9EbG1zmUDhBPGxTXp\nU/WVks5T5ajiZVV2anMHZoqIP0jfO31fUvV3HRsjYmb6zulMSUtU2UEP5xuqDCF+23ZPCr5JVffv\nqZp+VtIkSZ22J9j+ku1nbL8iaXeap3M0hdo+N/3A44X0+C+M9rHDmCfpz1JA/1zSS5Is6ay0zN9L\nQ34H0v0dDVhm9eszT9K1A8tPy7hM0twU7r+hypFpr+2/sX1encsGCCeMn4j4kaR7JP1J2qk9qspQ\n1ck8xz5VhgJrDalVz3c0Ij4XEReoMtT0AVWG8gacUzX9TlWOtPol/SdJKyW9V5Wd/Pw0jweeeoQu\n3inpZ5IWRcQMVYbjPPxDRrRH0idTQA9cTo2If07fL/1XSR+RNCuF+4ER+ntI0turbp9ZY57qx+1R\n5UizevlTI+JLkhQRD0fEf1Tlg8bPJH21nmIBSXIO/zKjs7Mz5s+f3+puoAmeeuopzZs3TzNmzJAk\nHT16VFu2bNHixYv1+uuva8eOHZo7d65OP/10TZo0SUeOHNHOnTs1Y8YMveMd71B/f7/6+/t13nmV\nD+PHjh3Tzp07NXHiRC1cuHDI5R48eFATJ07UlClTdPz4cW3fvl1z5sxRZ2ennn76aR0+fFiLFi3S\n5MmTtWvXLtnWwoULtX//fvX392vx4sqo2PPPP6++vj4tWbJEU6ZM0d69e3XkyJE3Lbu6xm3btqmj\no0Nz587V4cOH1d3drYkTJ57o/+bNm08811B6enp0+PBhLViwQJL08ssvq6enRwsXLtSpp56q48eP\n65VXXtGsWbN04MABPfvsszr//PM1YcIEvfDCC+rt7dWiRYs0Y8YM9fX16cUXX9TixYtlV/Jq165d\niggtWLBAr732mnbs2KGOjg4tWLBAhw8f1pYtW7R06dIT8x85ckTbtm3T/PnzNWPGDEWEDh06pMmT\nJ8u2Dh06pOnTp+uUU05Rb2+vDh48eOL1AwbbvHlzf0TMHnHGiGj55d3vfnegnObNmxcbNmx4U9un\nPvWp+NCHPhQRERs3bozly5dHR0dHdHR0xJIlS+KWW26J/v7+iIj4+te/HqecckpMnTo1pk6dGrNn\nz45Vq1bFvn37hl3ufffdF+eee268/e1vjzlz5sTNN98cR48ejYiIK664ItauXRvvec97Yvr06fGB\nD3wg+vr6IiLi4MGD8cEPfjCmTZsW73znO2PdunUhKXbs2BEREdu3b48LL7wwOjo6YuXKlW+p8Uc/\n+lEsXrw4pk6dGpdddll89rOfjWXLlp3oV/VzDeXWW2+N66677k1t9957b7zrXe+K6dOnx9lnnx3X\nX399REQcO3Ysbrjhhpg+fXqceeaZ8eUvf/lN/env749ly5bFzJkz4+KLL46IiGeeeSYuueSSmDp1\naqxYsSJuvvnmE8vbtWtXSDrxWg3YuHFjXH755TFr1qzo7OyMFStWxLPPPhs9PT1x+eWXx4wZM6Kj\noyOuuOKK2Lp167D1ob1J2hSjyIUsjpy6urpi06aRfnwFNMaVV16pj33sY/rEJz7R6q4Abcf25ojo\nGmk+vnMCAGSHcEJhrV+/XtOmTXvLZcmSJa3u2oiWLFlSs+/r169vddeALDCsBwAYNwzrAQAKi3AC\nAGSHcAIAZIdwAgBkh3ACAGSHcAIAZIdwAgBkh3ACAGSHcAIAZIdwAgBkh3ACAGSHcAIAZIdwAgBk\nh3ACAGSHcAIAZIdwAgBkh3ACAGSHcAIAZIdwAgBkh3ACAGSHcAIAZGfEcLJ9ju0f2N5me6vt30nt\np9neYHtHup6V2m37Dtvdtp+0vbTZRQAAymU0R07HJP1eRJwv6VJJN9m+QNJaSY9ExCJJj6TbkrRc\n0qJ0WSPpzob3GgAKxm51D4plxHCKiN6I+GmaPihpm6SzJK2UtC7Ntk7SNWl6paR7o2KjpJm25za8\n5wCA0jqp75xsz5d0saRHJZ0REb1SJcAkzUmznSVpT9XD9qa2wc+1xvYm25v6+vpOvucAgNIadTjZ\nnibpe5J+NyJeGW7WGm3xloaIuyKiKyK6Zs+ePdpuAADawKjCyfYkVYJpfUT8VWreNzBcl673p/a9\nks6pevjZknoa010AQDsYza/1LOlrkrZFxJ9W3fWgpNVperWkB6raP55+tXeppAMDw38AAIzGxFHM\ns0zSf5b0lO3HU9stkr4k6X7bN0p6TtK16b6HJK2Q1C3pNUnXN7THAIDSGzGcIuKfVPt7JEm6qsb8\nIemmOvsFAGhjnCECAJAdwgkAkB3CCQCQHcIJpcIpYoByIJwAoKDK/GGMcAIAZIdwqqHMn0YAoAgI\nJ6AJ+IAD1IdwAgBkh3AC2kwzjuo4UkSjEU4AgOwQTkALccRRfKzD5iCcAGAQAqf1CCcAQHYIJwBA\ndggnAG2D4briIJwAANkhnAAA2SGcAADZabtwYswZAPLXduEEAMgf4YSW42gWwGCEEwAgO6UKp/H8\nBM6nfdRjtNsP2xnaVanCCQBQDoQTACA7hNMwGFLJB+sCaC+EEwqDgALaB+EEAMgO4QQAyE7bhBND\nQkA58d4up7YJJwBAcRBOg/ApDABar5ThZBMyAFBkpQsnQgnIF+9PjFbpwgkYC3aaQF4IJwBZ4gND\neyOc0DTsXACMFeEEAMgO4YTS4YgNKD7CCS1FkACopW3D6WR2iuxAUQRspyiTtg0nYCwIAGB8EE7I\nAjt9ANUIJwBAdkYMJ9t3295ve0tV2222n7f9eLqsqLrvM7a7bT9t+/3N6jiKgSMiAGMxmiOneyRd\nXaP9v0fERenykCTZvkDSKklL0mP+p+0JjepsvdhRAkAxjBhOEfFjSS+N8vlWSvp2RByOiF2SuiVd\nUkf/MM4IcKD1eB/W953Tp20/mYb9ZqW2syTtqZpnb2p7C9trbG+yvamvr6+ObtRnqI2AjQMAWmes\n4XSnpF+SdJGkXklfSe21dulR6wki4q6I6IqIrtmzZ4+xG603mhAj6FqH1x4opjGFU0Tsi4jjEfG6\npK/qjaG7vZLOqZr1bEk99XURACGLdjOmcLI9t+rmr0sa+CXfg5JW2Z5se4GkRZIeq6+LQDGULUDK\nVk87K+K6nDjSDLa/JelKSZ2290q6VdKVti9SZchut6RPSlJEbLV9v6R/lXRM0k0Rcbw5XQcAlNWI\n4RQRH63R/LVh5v+8pM/X06mxaPYnA1uKmt+elU871QogT5whAgCQnbYOpyKOw+bA5rUD0FxtHU4A\ngDwRTmhLHP0BeSOcAADZIZwAANkhnDLDUBMAEE4YAiEJFFcZ3r+EEwqpDG8+oBb+U0IF4YSmGOsb\nKec3YM59A8qGcAIajBBDs7TTtkU4AQCyQzgBALJDOGWoGYfuA8853HO305ABgLwRTiUw2lAhfAAU\nBeEEYFzxIQmjQTgBALJDONUhp0+AOfUFAOpFOGHMCETkYjTbIttrsRBOAFqCsMBwCKcR5P4Gyr1/\nUjH6CCAvbRlOjd5ZNvPvksqME1wCGEpbhhNQFgQ5qpVpeyCcMjJ4wyrThtYoRTpKHel523n9lqX2\nHOvIsU9jQTgVWA4bYSPPTpFDPUXFa4eyIZwwpHbb4bVbvfXi9UIzEU4AgOwQTgAahqMpNArh1CC8\nKTFaRdpWitRXlAvhNE5s3ugoJrbbxuG1HD3CCYXFG728qtdtPeuZbWRoub82hBNKL/c3YSPxjyeL\nhz+zqI1wgqT23PjLYqzrbrzWeW7bVm79QW2EU5PxRigf1mnrsQ7qM/D65XzURjgB44idavthnY8N\n4QQMg+9wGqNV5y9s9vLLrvoIa7xfQ8IJJ4U3eeMV+TXNqe859QX1I5waLMc3SLP7NN415/gaN1NZ\n6m1UHSfzfQmKi3DKXJH+RUSuy8UbWAf1K9JrWKS+DkY4AW2iyDuqHBXl/3Xl0o+TRTgBALJDOAFo\niKJ+Qi+idnitCacGaocNBhhPrXxP8X5uLcIJQF1y34nn8KOiVvQh9/UyEsKpBYq+0aA12G4gtc92\nMGI42b7b9n7bW6raTrO9wfaOdD0rtdv2Hba7bT9pe2kzOw/grYq68yrC3y9xRorxM5ojp3skXT2o\nba2kRyJikaRH0m1JWi5pUbqskXRnY7oJNAc7ESBPI4ZTRPxY0kuDmldKWpem10m6pqr93qjYKGmm\n7bmN6izQCgQYxoptZ+zG+p3TGRHRK0npek5qP0vSnqr59qa2t7C9xvYm25v6+vrG2A2gtqLvFIre\nf6Bejf5BRK23VNSaMSLuioiuiOiaPXt2g7uRr0afXyy350I5sE20n9zW+VjDad/AcF263p/a90o6\np2q+syX1jL17KKLcNnIAw8vxPTvWcHpQ0uo0vVrSA1XtH0+/2rtU0oGB4T8AyF0r/m9R9bLxhokj\nzWD7W5KulNRpe6+kWyV9SdL9tm+U9Jyka9PsD0laIalb0muSrm9Cn0vBlqLmgCfQvthBY8CI4RQR\nHx3irqtqzBuSbqq3UwCA9sYZIgAA2SGcALxJuw6ttaLuwcts19e+FsIJQNsjFCpyeh0IJyBTOe0o\niojXr9gIJ5QWOyeguAinEmPnDKCoCCdgFAh6lFWu2zbhBAAtlmtAtBLhhFHjDQRgvBBOAIDsEE4A\ngOwQTgCA7BBOAIDsEE4AgOwQTgCA7BBOAIDsEE4lx98mtSfWO4qOcAIAZIdwAjCikzkS46gNjUA4\nAQCyQzgBVfjUD+SBcAIAZIdwAgBkh3ACCqTRw44MYyJXhFMBsUMBUHaEEwAgO4QTACA7hBMwBgyt\nAs1FOAEAskM4ARngSAx4M8IJAJAdwgkAkB3CCWgihuuAsSGcAADZIZwAANkhnAAA2SGcAADZIZwA\nANkhnAAA2SGcAADZIZwAANkhnAAA2SGcAADZIZwAANkhnIAG4lx6QGNMrOfBtndLOijpuKRjEdFl\n+zRJ35E0X9JuSR+JiJfr6yYAoJ004sjp30fERRHRlW6vlfRIRCyS9Ei6DQDAqDVjWG+lpHVpep2k\na5qwDABAidUbTiHp721vtr0mtZ0REb2SlK7n1Hqg7TW2N9ne1NfXV2c3AABlUtd3TpKWRUSP7TmS\nNtj+2WgfGBF3SbpLkrq6uqLOfgAASqSuI6eI6EnX+yV9X9IlkvbZnitJ6Xp/vZ0EALSXMYeT7am2\npw9MS3qfpC2SHpS0Os22WtID9XYSANBe6hnWO0PS9135w46Jku6LiL+z/RNJ99u+UdJzkq6tv5sA\ngHYy5nCKiJ2SLqzR/qKkq+rpFACgvXGGCABAdggnAEB2CCcAQHYIJwBAdggnAEB2CCcAQHYIJwBA\ndggnAEB2CCcAQHYIJwBAdggnAEB2CCcAQHYIJwBAdggnAEB2CCcAQHYIJwBAdggnAEB2CCcAQHYI\nJwBAdggnAEB2CCcAQHYIJwBAdggnAEB2CCcAQHYIJwBAdggnAEB2CCcAQHYIJwBAdggnAEB2CCcA\nQHYIJwBAdggnAEB2CCcAQHYIJwBAdggnAEB2CCcAQHYIJwBAdggnAEB2CCcAQHYIJwBAdggnAEB2\nCCcAQHYIJwBAdggnAEB2CCcAQHaaFk62r7b9tO1u22ubtRwAQPk0JZxsT5D055KWS7pA0kdtX9CM\nZQEAyqdZR06XSOqOiJ0RcUTStyWtbNKyAAAlM7FJz3uWpD1Vt/dK+jfVM9heI2lNuvmq7afrXGan\npP43L2N0DyzQfG+pcSzPN8Zlj8d8J+prVf+avOyGrL+M52v5+mvyfEOuv0y2r0bM12k3Zh8zjHmj\nmalZ4VSrhHjTjYi7JN3VsAXamyKiq1HPl6Oy10h9xUZ9xZdTjc0a1tsr6Zyq22dL6mnSsgAAJdOs\ncPqJpEW2F9h+m6RVkh5s0rIAACXTlGG9iDhm+9OSHpY0QdLdEbG1Gcuq0rAhwoyVvUbqKzbqK75s\nanREjDwXAADjiDNEAACyQzgBALJTinAqy6mSbO+2/ZTtx21vSm2n2d5ge0e6npXabfuOVPOTtpe2\ntvdvZftu2/ttb6lqO+l6bK9O8++wvboVtQxliBpvs/18Wo+P215Rdd9nUo1P235/VXt227Dtc2z/\nwPY221tt/05qL8U6HKa+Uqw/SbI9xfZjtp9INX4utS+w/WhaH99JP1yT7cnpdne6f37Vc9WsvWki\notAXVX5w8YykhZLeJukJSRe0ul9jrGW3pM5BbbdLWpum10r6cppeIelvVfmbskslPdrq/teo53JJ\nSyVtGWs9kk6TtDNdz0rTs1pd2wg13ibp92vMe0HaPidLWpC22wm5bsOS5kpamqanS9qeaijFOhym\nvlKsv9RnS5qWpidJejStm/slrUrtfyHpt9L0b0v6izS9StJ3hqu9mX0vw5FT2U+VtFLSujS9TtI1\nVe33RsVGSTNtz21FB4cSET+W9NKg5pOt5/2SNkTESxHxsqQNkq5ufu9HZ4gah7JS0rcj4nBE7JLU\nrcr2m+U2HBG9EfHTNH1Q0jZVzv5SinU4TH1DKdT6k6S0Ll5NNyelS0j6D5L+MrUPXocD6/YvJV1l\n2xq69qYpQzjVOlXScBtYzkLS39ve7MrpnSTpjIjolSpvJklzUntR6z7Zeopa56fT0NbdA8NeKnCN\naXjnYlU+eZduHQ6qTyrR+rM9wfbjkvar8sHgGUk/j4hjaZbq/p6oJd1/QNLpakGNZQinEU+VVCDL\nImKpKmdzv8n25cPMW6a6paHrKWKdd0r6JUkXSeqV9JXUXsgabU+T9D1JvxsRrww3a422ItZXqvUX\nEccj4iJVztRziaTza82WrrOpsQzhVJpTJUVET7reL+n7qmxI+waG69L1/jR7Ues+2XoKV2dE7Es7\nhNclfVVvDH8Urkbbk1TZca+PiL9KzaVZh7XqK9P6qxYRP5f0Q1W+c5ppe+AkDNX9PVFLur9DlWHr\nca+xDOFUilMl2Z5qe/rAtKT3SdqiSi0Dv25aLemBNP2gpI+nX0hdKunAwFBL5k62noclvc/2rDS8\n8r7Ulq1B3/39uirrUarUuCr9ImqBpEWSHlOm23D6ruFrkrZFxJ9W3VWKdThUfWVZf5Jke7btmWn6\nVEnvVeW7tR9I+nCabfA6HFi3H5b0D1H5RcRQtTfPePxipNkXVX4ltF2VsdQ/bHV/xljDQlV+DfOE\npK0Ddagy3vuIpB3p+rR441c4f55qfkpSV6trqFHTt1QZFjmqyievG8dSj6QbVPkCtlvS9a2uaxQ1\nfiPV8KQqb+q5VfP/YarxaUnLc96GJV2mytDNk5IeT5cVZVmHw9RXivWX+vUrkv5fqmWLpD9K7QtV\nCZduSd+VNDm1T0m3u9P9C0eqvVkXTl8EAMhOGYb1AAAlQzgBALJDOAEAskM4AQCyQzgBALJDOAEA\nskM4AQCy8/8BpsaziGw3kCsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f8695ee4b38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAawAAAEdCAYAAABQXlN8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJEmd8P8Bc58FU5ludrIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f86939bd390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaQAAAEdCAYAAABDiROIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJh+r23sAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f869700c940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAARoAAAEdCAYAAADTtqgCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAIqDowGA4uBoAKA4/w+AlTqT3c57ewAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f8695e25a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAARoAAAEdCAYAAADTtqgCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f86974be240>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAARoAAAEdCAYAAADTtqgCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAIrzfwHtfosv1DswAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f869678e7b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "array([  1.46000000e+02,   1.56000000e+02,   1.37000000e+02, ...,\n",
       "         1.53337504e-02,   1.80869185e-02,   6.85306287e-03])"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the profile time for a single sample\n",
    "# t=time.time()        \n",
    "file='test_images/not_car.png'\n",
    "# file='test_images/car.png'\n",
    "# image = cv2.imread(file)\n",
    "image = scipy.misc.imread(file)\n",
    "size = (64,64)\n",
    "img = cv2.resize(image, size)#.astype(np.float32)\n",
    "extract_features_img(img,cspace='RGB',debug=1)\n",
    "# extract_features_img(img,cspace='HSV',debug=1)\n",
    "# extract_features_img(img,cspace='LUV',debug=1)\n",
    "# extract_features_img(img,cspace='HLS',debug=1)\n",
    "# extract_features_img(img,cspace='YUV',debug=1)\n",
    "# extract_features_img(img,cspace='YCrCb',debug=1)\n",
    "# t2 = time.time()\n",
    "# print(round(t2-t,1), 'Seconds')"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda env:carnd-term1]",
   "language": "python",
   "name": "conda-env-carnd-term1-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}