supplemental_image_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "702233db",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "import numpy as np\n",
    "from scipy.optimize import linear_sum_assignment\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ca6cf011",
   "metadata": {},
   "outputs": [],
   "source": [
    "def histogram_intersection(p, q):\n",
    "    # Calculate histogram intersection\n",
    "    # q is the reference distribution\n",
    "    # p is the query distribution\n",
    "    \n",
    "    minima = np.minimum(p, q)\n",
    "    hi = np.true_divide(np.sum(minima), np.sum(p))\n",
    "    \n",
    "    if hi > 1 or hi < 0:\n",
    "        print('Error, HI =', hi)\n",
    "        print(p)\n",
    "        print(q)\n",
    "        return\n",
    "    return hi\n",
    "\n",
    "\n",
    "def chi_square_distance(p, q):\n",
    "    # Calculate chi-square distance\n",
    "    # q is the reference distribution\n",
    "    # p is the query distribution\n",
    "    \n",
    "    p = np.array(p)/np.sum(p)\n",
    "    q = np.array(q)/np.sum(q)\n",
    "    \n",
    "    p = np.array(p) + 1\n",
    "    q = np.array(q) + 1\n",
    "    \n",
    "    return np.sum(((p - q)**2 / (p + q))) / 2\n",
    "\n",
    "\n",
    "def kl_divergence(p, q):\n",
    "    # Calculate Kullback-Leibler Divergence\n",
    "    # q is the reference distribution\n",
    "    # p is the query distribution\n",
    "    \n",
    "    p = np.array(p, dtype=float)\n",
    "    q = np.array(q, dtype=float)\n",
    "    \n",
    "    p = np.array(p) + 1\n",
    "    q = np.array(q) + 1\n",
    "    \n",
    "    p /= p.sum()\n",
    "    q /= q.sum()\n",
    "    \n",
    "    kl_div = np.sum(p * np.log(p / q))\n",
    "    return kl_div\n",
    "\n",
    "\n",
    "def earth_movers_distance(p, q):\n",
    "    # Calculate Earth Mover's Distance\n",
    "    # q is the reference distribution\n",
    "    # p is the query distribution\n",
    "    \n",
    "    p = np.array(p, dtype=float)\n",
    "    q = np.array(q, dtype=float)\n",
    "\n",
    "    # Normalize the distributions to ensure they sum to 1\n",
    "    p /= p.sum()\n",
    "    q /= q.sum()\n",
    "\n",
    "    # Calculate cumulative distributions\n",
    "    P = np.cumsum(p)\n",
    "    Q = np.cumsum(q)\n",
    "\n",
    "    # Calculate the cost matrix\n",
    "    C = np.abs(np.subtract.outer(P, Q))\n",
    "\n",
    "    # Solve linear sum assignment problem\n",
    "    row_ind, col_ind = linear_sum_assignment(C)\n",
    "    emd = C[row_ind, col_ind].sum()\n",
    "\n",
    "    return emd\n",
    "\n",
    "\n",
    "def kolmogorov_smirnov_distance(p, q):\n",
    "    \n",
    "    # Calculate KS distance\n",
    "    # q is the reference distribution\n",
    "    # p is the query distribution\n",
    "    \n",
    "    p = np.array(p, dtype=float)\n",
    "    q = np.array(q, dtype=float)\n",
    "\n",
    "    # Normalize the distributions to ensure they sum to 1\n",
    "    p /= p.sum()\n",
    "    q /= q.sum()\n",
    "\n",
    "    # Calculate cumulative distributions\n",
    "    P = np.cumsum(p)\n",
    "    Q = np.cumsum(q)\n",
    "\n",
    "    ks_distance = np.max(np.abs(P - Q))\n",
    "\n",
    "    return ks_distance\n",
    "\n",
    "\n",
    "def rank_probability_score(p, q):\n",
    "    \n",
    "    # Calculate the ranked probability score\n",
    "    # q is the reference distribution\n",
    "    # p is the query distribution\n",
    "    \n",
    "    p = np.array(p, dtype=float)\n",
    "    q = np.array(q, dtype=float)\n",
    "\n",
    "    # Normalize the distributions to ensure they sum to 1\n",
    "    p /= p.sum()\n",
    "    q /= q.sum()\n",
    "\n",
    "    # Calculate cumulative distributions\n",
    "    P = np.cumsum(p)\n",
    "    Q = np.cumsum(q)\n",
    "\n",
    "    rps = np.sum((P - Q)**2)\n",
    "\n",
    "    return rps\n",
    "\n",
    "\n",
    "def RDS(p, q):\n",
    "    \n",
    "    # Calculate Relative distribution shift\n",
    "    # q is the reference distribution\n",
    "    # p is the query distribution\n",
    "    \n",
    "    p_bins = len(p)\n",
    "    p_obs = sum(p)\n",
    "    \n",
    "    q_bins = len(q)\n",
    "    q_obs = sum(q)\n",
    "    \n",
    "    z_p = (p_bins + 1)/p_bins\n",
    "    p = [sum(p[:ii+1])**(z_p) for ii in range(len(p))]\n",
    "    Sp = np.sum(np.array(p)/(p_obs**z_p)) - 1\n",
    "    Sp = Sp/(p_bins - 1)\n",
    "    \n",
    "    z_q = (q_bins + 1)/q_bins\n",
    "    q = [sum(q[:ii+1])**(z_q) for ii in range(len(q))]\n",
    "    Sq = np.sum(np.array(q)/(q_obs**z_q)) - 1\n",
    "    Sq = Sq/(q_bins - 1)\n",
    "    \n",
    "    return Sq - Sp\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9450cdef",
   "metadata": {},
   "outputs": [],
   "source": [
    "def flatten_histogram(image, num_bins, htype='color'):\n",
    "    \n",
    "    if htype == 'color':\n",
    "        \"\"\"\n",
    "        Construct a flattened 1D histogram for a color image by concatenating the histograms\n",
    "        of each color channel.\n",
    "        \"\"\"\n",
    "        if len(image.shape) == 3:  # Check if image is RGB\n",
    "            channels = cv2.split(image)\n",
    "            histograms = []\n",
    "            for chan in channels:\n",
    "                hist, _ = np.histogram(chan, bins=num_bins, range=(0, num_bins))\n",
    "                histograms.append(hist)\n",
    "            histogram = np.concatenate(histograms) / np.sum(histograms)\n",
    "            \n",
    "        else:  # For grayscale images, compute histogram directly\n",
    "            histogram, _ = np.histogram(image, bins=num_bins, range=(0, num_bins))\n",
    "            histogram = histogram / np.sum(histogram)\n",
    "        return histogram\n",
    "    \n",
    "    elif htype == 'hue':\n",
    "        \"\"\"\n",
    "        Construct a 1D hue histogram for a color image by converting it to the HSV color space\n",
    "        and calculating the histogram of the hue component.\n",
    "        \"\"\"\n",
    "        # Convert RGB image to HSV\n",
    "        hsv_image = cv2.cvtColor(image, cv2.COLOR_RGB2HSV)\n",
    "\n",
    "        # Extract Hue component\n",
    "        hue_channel = hsv_image[:, :, 0]\n",
    "\n",
    "        # Calculate histogram of the Hue component\n",
    "        histogram, _ = np.histogram(hue_channel, bins=num_bins, range=(0, 180))  # Hue values range from 0 to 180 in OpenCV\n",
    "\n",
    "        # Normalize histogram\n",
    "        histogram = histogram / np.sum(histogram)\n",
    "    \n",
    "        return histogram\n",
    "    \n",
    "    \n",
    "    elif htype == 'intensity':\n",
    "        \"\"\"\n",
    "        Construct a 1D intensity histogram for an image by converting it to grayscale\n",
    "        and calculating the histogram based on the intensity values.\n",
    "        \"\"\"\n",
    "        # Check if image is in grayscale\n",
    "        if len(image.shape) == 3:  # If not, convert image to grayscale\n",
    "            grayscale_image = cv2.cvtColor(image, cv2.COLOR_RGB2GRAY)\n",
    "        else:\n",
    "            grayscale_image = image  # The image is already in grayscale\n",
    "\n",
    "        # Calculate the histogram of intensity values\n",
    "        histogram, _ = np.histogram(grayscale_image, bins=num_bins, range=(0, 256))\n",
    "\n",
    "        # Normalize\n",
    "        histogram = histogram / np.sum(histogram)\n",
    "        return histogram\n",
    "    \n",
    "\n",
    "\n",
    "\n",
    "def process_video_with_metric(video_path, metric_name, num_bins):\n",
    "    video = cv2.VideoCapture(video_path)\n",
    "    total_frames = int(video.get(cv2.CAP_PROP_FRAME_COUNT))\n",
    "    metric_values = []\n",
    "    frame_indices = []\n",
    "    first_frame = None\n",
    "    max_rds_value = None\n",
    "    max_rds_frame = None\n",
    "\n",
    "    # Define videos that need trimming\n",
    "    videos_to_trim = ['data/time_lapse_video/855852-sd_640_360_30fps.mp4', \n",
    "                      'data/time_lapse_video/856030-sd_640_360_25fps.mp4',\n",
    "                      'data/time_lapse_video/856006-sd_640_360_25fps.mp4',\n",
    "                      'data/time_lapse_video/856990-sd_640_360_30fps.mp4',\n",
    "                      'data/time_lapse_video/2038351-sd_640_360_30fps.mp4',\n",
    "                      'data/time_lapse_video/855215-sd_640_360_24fps.mp4',\n",
    "                      'data/time_lapse_video/3541930-sd_426_240_24fps.mp4',\n",
    "                      'data/time_lapse_video/9255202-sd_426_240_25fps.mp4',\n",
    "                      'data/time_lapse_video/9255065-sd_426_240_25fps.mp4',\n",
    "                      'data/time_lapse_video/10881636-sd_640_360_25fps.mp4',\n",
    "                      'data/time_lapse_video/10881635-sd_640_360_25fps.mp4',\n",
    "                      'data/time_lapse_video/10881640-sd_640_360_25fps.mp4',\n",
    "                      'data/time_lapse_video/14668-258508783_tiny.mp4',\n",
    "                      'data/time_lapse_video/11440-230272102_tiny.mp4',\n",
    "                      'data/time_lapse_video/9289-218679635_tiny.mp4',\n",
    "                      'data/time_lapse_video/16986-277920487_tiny.mp4',\n",
    "                      'data/time_lapse_video/8560-211218186_tiny.mp4',\n",
    "                      'data/time_lapse_video/18020-287831424_tiny.mp4',\n",
    "                      'data/time_lapse_video/81089-574743647_medium.mp4',\n",
    "                      \n",
    "                 ]\n",
    "\n",
    "    # Determine if current video needs trimming\n",
    "    trim = video_path in videos_to_trim\n",
    "\n",
    "    # Adjust start and end indices if trimming is needed\n",
    "    start_index = 10 if trim else 0\n",
    "    end_index = total_frames - 10 if trim else total_frames\n",
    "\n",
    "    for frame_count in range(total_frames):\n",
    "        success, frame = video.read()\n",
    "        if not success:\n",
    "            break\n",
    "\n",
    "        # Skip processing for trimmed frames\n",
    "        if frame_count < start_index or frame_count >= end_index:\n",
    "            continue\n",
    "\n",
    "        if metric_name == 'RDS' and frame_count == start_index:\n",
    "            first_frame = frame  # Adjusted to capture the first untrimmed frame\n",
    "\n",
    "        if frame_count == start_index:  # Adjusted to use the first untrimmed frame as reference\n",
    "            reference_hist = flatten_histogram(frame, num_bins, 'color')\n",
    "\n",
    "        frame_hist = flatten_histogram(frame, num_bins, 'color')\n",
    "\n",
    "        if metric_name == '1 - HI':\n",
    "            metric_value = 1 - histogram_intersection(frame_hist, reference_hist)\n",
    "        elif metric_name == 'Chi-Square':\n",
    "            metric_value = chi_square_distance(frame_hist, reference_hist)\n",
    "        elif metric_name == 'KL':\n",
    "            metric_value = kl_divergence(frame_hist, reference_hist)\n",
    "        elif metric_name == 'EMD':\n",
    "            metric_value = earth_movers_distance(frame_hist, reference_hist)\n",
    "        elif metric_name == 'KS':\n",
    "            metric_value = kolmogorov_smirnov_distance(frame_hist, reference_hist)\n",
    "        elif metric_name == 'RPS':\n",
    "            metric_value = rank_probability_score(frame_hist, reference_hist)\n",
    "        elif metric_name == 'RDS':\n",
    "            metric_value = RDS(frame_hist, reference_hist)\n",
    "\n",
    "        metric_values.append(metric_value)\n",
    "\n",
    "        # Update max RDS frame and value if this is the RDS metric\n",
    "        if metric_name == 'RDS' and (max_rds_value is None or abs(metric_value) > abs(max_rds_value)):\n",
    "            max_rds_value = metric_value\n",
    "            max_rds_frame = frame\n",
    "\n",
    "        frame_indices.append(frame_count)\n",
    "\n",
    "    video.release()\n",
    "    return metric_values, frame_indices, first_frame, max_rds_frame, max_rds_value\n",
    "\n",
    "\n",
    "\n",
    "def make_fig(inputs, lab, first_frame, max_rds_frame, max_rds_value):\n",
    "    fs = 18\n",
    "    fig = plt.figure(figsize=(18, 18))  # Adjusted size for 3x3 grid\n",
    "    \n",
    "    # Function to rotate an image if width > height\n",
    "    def rotate_if_needed(image):\n",
    "        if image.shape[1] > image.shape[0]:  # Check if width > height\n",
    "            return cv2.rotate(image, cv2.ROTATE_90_CLOCKWISE)  # Rotate 90 degrees clockwise\n",
    "        return image\n",
    "\n",
    "    # Apply rotation if needed and plot the first frame\n",
    "    rotated_first_frame = rotate_if_needed(first_frame)\n",
    "    ax1 = plt.subplot(3, 3, 1)\n",
    "    ax1.imshow(cv2.cvtColor(rotated_first_frame, cv2.COLOR_BGR2RGB))\n",
    "    ax1.set_title('First frame', fontsize=fs-2)\n",
    "    ax1.axis('off')\n",
    "\n",
    "    # Apply rotation if needed and plot the frame with the greatest absolute RDS value\n",
    "    rotated_max_rds_frame = rotate_if_needed(max_rds_frame)\n",
    "    ax2 = plt.subplot(3, 3, 2)\n",
    "    ax2.imshow(cv2.cvtColor(rotated_max_rds_frame, cv2.COLOR_BGR2RGB))\n",
    "    ax2.set_title(f'RDS: {max_rds_value:.3f}', fontsize=fs-2)\n",
    "    ax2.axis('off')\n",
    "\n",
    "    # Plot the existing metric plots in the rest of the grid\n",
    "    for i, ls in enumerate(inputs, start=2):\n",
    "        ax = plt.subplot(3, 3, i+1)\n",
    "        plt.plot(ls[1], ls[0], c='k')\n",
    "        plt.xlabel('Frame', fontsize=fs)\n",
    "        plt.ylabel(ls[2], fontsize=fs, fontweight='bold')\n",
    "        plt.tick_params(axis='both', labelsize=10)\n",
    "\n",
    "    fig.patch.set_facecolor('white')\n",
    "    plt.subplots_adjust(hspace=0.35, wspace=0.4)\n",
    "    plt.savefig('Final_Figs/from_image_analysis/'+lab+'.pdf', bbox_inches='tight', format='pdf', dpi=600)\n",
    "    plt.close()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2b9365a9",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Already exists: 3541930-sd_426_240_24fps\n",
      "Already exists: 2559431-sd_426_240_24fps\n",
      "Already exists: 1110140-sd_640_360_30fps\n",
      "Already exists: 3053992-sd_426_240_30fps\n",
      "Already exists: 988047-sd_640_360_30fps\n",
      "Already exists: 5595352-sd_240_426_24fps\n",
      "Already exists: 856136-sd_640_360_25fps\n",
      "Already exists: 857251-sd_540_360_25fps\n",
      "Already exists: 2675512-sd_426_240_24fps\n",
      "Already exists: 855456-sd_640_360_30fps\n",
      "Already exists: 855210-sd_640_360_30fps\n",
      "Already exists: 2038351-sd_640_360_30fps\n",
      "Already exists: 855215-sd_640_360_24fps\n",
      "Already exists: 855852-sd_640_360_30fps\n",
      "Already exists: 856990-sd_640_360_30fps\n",
      "Already exists: 1779441-sd_640_360_30fps\n",
      "Already exists: 856006-sd_640_360_25fps\n",
      "Already exists: 856030-sd_640_360_25fps\n",
      "Already exists: 9837842-sd_960_540_30fps\n",
      "Already exists: 10661569-sd_426_240_30fps\n",
      "Already exists: 2731905-hd_1280_720_24fps\n",
      "Already exists: 4782177-hd_1280_720_30fps\n",
      "Already exists: 5158727-hd_1280_720_30fps\n",
      "Already exists: 14219843-sd_426_240_24fps\n",
      "Already exists: 14219844-sd_426_240_24fps\n",
      "Already exists: 3947230-sd_426_240_25fps\n",
      "Already exists: 10881640-sd_640_360_25fps\n",
      "Already exists: 10881635-sd_640_360_25fps\n",
      "Already exists: 10881636-sd_640_360_25fps\n",
      "Already exists: 4990234-sd_426_240_30fps\n",
      "Already exists: 6923433-sd_426_240_30fps\n",
      "Already exists: 4990232-sd_426_240_30fps\n",
      "Already exists: 4990233-sd_426_240_30fps\n",
      "Already exists: 9255202-sd_426_240_25fps\n",
      "Already exists: 7235032-sd_240_426_30fps\n",
      "Already exists: 7235041-sd_240_426_30fps\n",
      "Already exists: 5388577-sd_240_426_25fps\n",
      "Already exists: 5094592-sd_226_426_25fps\n",
      "Already exists: 7101966-hd_1080_1644_25fps\n",
      "Already exists: 17333461-sd_426_240_25fps\n",
      "Already exists: 4943067-sd_426_240_24fps\n",
      "Already exists: 7710243-sd_426_240_30fps\n",
      "Already exists: 10613972-hd_1920_1080_24fps\n",
      "Already exists: 10413970-hd_1920_1080_30fps\n",
      "Already exists: 7235033-sd_240_426_30fps\n",
      "Already exists: 10469592-sd_640_360_30fps\n",
      "Already exists: 4584697-hd_1920_1080_25fps\n",
      "Already exists: 7703989-hd_1080_1920_25fps\n",
      "Already exists: 7234900-hd_1920_1080_30fps\n",
      "Already exists: 5094592-hd_1080_2048_25fps\n",
      "Already exists: 7565623-hd_1080_1920_25fps\n",
      "Already exists: 5926164-hd_1920_1080_30fps\n",
      "Already exists: 8516544-hd_1080_1920_30fps\n",
      "Already exists: 4378289-hd_1920_1080_24fps\n",
      "Already exists: 1681023-sd_640_338_24fps\n",
      "Already exists: 81092-574743664_medium\n",
      "Already exists: 81089-574743647_medium\n",
      "Already exists: 12902-242487547_small\n",
      "Already exists: 1378-147055474_medium\n",
      "Already exists: 21161-315224572_medium\n",
      "Already exists: 72512-543388307_medium\n",
      "Already exists: 8782-214159393_tiny\n",
      "Already exists: 7684-202716554_medium\n",
      "Already exists: 91592-628463727_medium\n",
      "Already exists: 18327-291012897_small\n",
      "Already exists: 188766-883827686_small\n",
      "Already exists: 55303-499594258_small\n",
      "Already exists: 90559-626984434_small\n",
      "Already exists: 12719-241674195_medium\n",
      "Already exists: 14124-254487707_tiny\n",
      "Already exists: 55554-502339972_small\n",
      "Already exists: 18324-291012885_small\n",
      "Already exists: 80645-572367655_medium\n",
      "Already exists: 18020-287831424_tiny\n",
      "Already exists: 75475-556034323_small\n",
      "Already exists: 24718-345209742_small\n",
      "Already exists: 37311-413624981_small\n",
      "Already exists: 8560-211218186_tiny\n",
      "Already exists: 16986-277920487_tiny\n",
      "Already exists: 151406-800680974_tiny\n",
      "Already exists: 144761-785265038_tiny\n",
      "['9289-218679635_tiny']\n",
      "['11440-230272102_tiny']\n",
      "['14668-258508783_tiny']\n",
      "['12668-241236737_tiny']\n",
      "['122211-724743609_tiny']\n",
      "['122336-725067675_tiny']\n",
      "['24715-345209726_small']\n",
      "['66262-516980798_tiny']\n",
      "['4935-181466751_medium']\n",
      "['43781-436252322_tiny']\n",
      "['18251-290360026_tiny']\n",
      "['55776-503981063_small']\n",
      "['18435-292228551_tiny']\n",
      "['18433-292228530_tiny']\n",
      "['81185-575775322_tiny']\n",
      "['1285-145117054_small']\n",
      "['6089-188582377_tiny']\n",
      "['15708-266043570_medium']\n",
      "['18902-297379468_tiny']\n",
      "['22352-326722819_medium']\n",
      "['25694-352026464_small']\n",
      "['40487-425442645_small']\n",
      "['43448-436041766_tiny']\n",
      "['122213-724743622_tiny']\n",
      "['124277-730508581_medium']\n",
      "['131398-750568980_tiny']\n",
      "['135422-761676527_small']\n",
      "['135578-762107314_medium']\n",
      "['166284-834580727_tiny']\n",
      "Already exists: 11440-230272102_tiny\n",
      "Already exists: 14668-258508783_tiny\n",
      "Already exists: 9289-218679635_tiny\n",
      "Already exists: 16986-277920487_tiny\n",
      "Already exists: 8560-211218186_tiny\n",
      "Already exists: 18020-287831424_tiny\n",
      "Already exists: 81089-574743647_medium\n"
     ]
    }
   ],
   "source": [
    "# From pexels\n",
    "pexel_lists = [['3541930-sd_426_240_24fps'],\n",
    "         ['2559431-sd_426_240_24fps'],\n",
    "         ['1110140-sd_640_360_30fps'],\n",
    "         ['3053992-sd_426_240_30fps'],\n",
    "         ['988047-sd_640_360_30fps'],\n",
    "         ['5595352-sd_240_426_24fps'],\n",
    "         ['856136-sd_640_360_25fps'],\n",
    "         ['857251-sd_540_360_25fps'],\n",
    "         ['2675512-sd_426_240_24fps'],\n",
    "         ['855456-sd_640_360_30fps'],\n",
    "         ['855210-sd_640_360_30fps'],\n",
    "         ['2038351-sd_640_360_30fps'],\n",
    "         ['855215-sd_640_360_24fps'],\n",
    "         ['855852-sd_640_360_30fps'],\n",
    "         ['856990-sd_640_360_30fps'],\n",
    "         ['1779441-sd_640_360_30fps'],\n",
    "         ['856006-sd_640_360_25fps'],\n",
    "         ['856030-sd_640_360_25fps'],\n",
    "         ['9837842-sd_960_540_30fps'],\n",
    "         ['10661569-sd_426_240_30fps'],\n",
    "         ['2731905-hd_1280_720_24fps'],\n",
    "         ['4782177-hd_1280_720_30fps'],\n",
    "         ['5158727-hd_1280_720_30fps'],\n",
    "         ['14219843-sd_426_240_24fps'],\n",
    "         ['14219844-sd_426_240_24fps'],\n",
    "         ['3947230-sd_426_240_25fps'],\n",
    "         ['10881640-sd_640_360_25fps'],\n",
    "         ['10881635-sd_640_360_25fps'],\n",
    "         ['10881636-sd_640_360_25fps'],\n",
    "         ['4990234-sd_426_240_30fps'],\n",
    "         ['6923433-sd_426_240_30fps'],\n",
    "         ['4990232-sd_426_240_30fps'],\n",
    "         ['4990233-sd_426_240_30fps'],\n",
    "         ['9255202-sd_426_240_25fps'],\n",
    "         ['7235032-sd_240_426_30fps'],\n",
    "         ['7235041-sd_240_426_30fps'],\n",
    "         ['5388577-sd_240_426_25fps'],\n",
    "         ['5094592-sd_226_426_25fps'],\n",
    "         ['7101966-hd_1080_1644_25fps'],\n",
    "         ['17333461-sd_426_240_25fps'],\n",
    "         ['4943067-sd_426_240_24fps'],\n",
    "         ['7710243-sd_426_240_30fps'],\n",
    "         ['10613972-hd_1920_1080_24fps'],\n",
    "         ['10413970-hd_1920_1080_30fps'],\n",
    "         ['7235033-sd_240_426_30fps'],\n",
    "         ['10469592-sd_640_360_30fps'],\n",
    "         ['4584697-hd_1920_1080_25fps'],\n",
    "         ['7703989-hd_1080_1920_25fps'],\n",
    "         ['7234900-hd_1920_1080_30fps'],\n",
    "         ['5094592-hd_1080_2048_25fps'],\n",
    "         ['7565623-hd_1080_1920_25fps'],\n",
    "         ['5926164-hd_1920_1080_30fps'],\n",
    "         ['8516544-hd_1080_1920_30fps'],\n",
    "         ['4378289-hd_1920_1080_24fps'],\n",
    "         ['1681023-sd_640_338_24fps'],\n",
    "        ]\n",
    "\n",
    "\n",
    "# From Pixabay\n",
    "pixabay_lists = [['81092-574743664_medium'],\n",
    "         ['81089-574743647_medium'],\n",
    "         ['12902-242487547_small'],\n",
    "         ['1378-147055474_medium'],\n",
    "         ['21161-315224572_medium'],\n",
    "         ['72512-543388307_medium'],\n",
    "         ['8782-214159393_tiny'],\n",
    "         ['7684-202716554_medium'],\n",
    "         ['91592-628463727_medium'],\n",
    "         ['18327-291012897_small'],\n",
    "         ['188766-883827686_small'],\n",
    "         ['55303-499594258_small'],\n",
    "         ['90559-626984434_small'],\n",
    "         ['12719-241674195_medium'],\n",
    "         ['14124-254487707_tiny'],\n",
    "         ['55554-502339972_small'],\n",
    "         ['18324-291012885_small'],\n",
    "         ['80645-572367655_medium'],\n",
    "         ['18020-287831424_tiny'],\n",
    "         ['75475-556034323_small'],\n",
    "         ['24718-345209742_small'],\n",
    "         ['37311-413624981_small'],\n",
    "         ['8560-211218186_tiny'],\n",
    "         ['16986-277920487_tiny'],\n",
    "         ['151406-800680974_tiny'],\n",
    "         ['144761-785265038_tiny'],\n",
    "         ['9289-218679635_tiny'],\n",
    "         ['11440-230272102_tiny'],\n",
    "         ['14668-258508783_tiny'],\n",
    "         ['12668-241236737_tiny'],\n",
    "         ['122211-724743609_tiny'],\n",
    "         ['122336-725067675_tiny'],\n",
    "         ['24715-345209726_small'],\n",
    "         ['66262-516980798_tiny'],\n",
    "         ['4935-181466751_medium'],\n",
    "         ['43781-436252322_tiny'],\n",
    "         ['18251-290360026_tiny'],\n",
    "         ['55776-503981063_small'],\n",
    "         ['18435-292228551_tiny'],\n",
    "         ['18433-292228530_tiny'],\n",
    "         ['81185-575775322_tiny'],\n",
    "         ['1285-145117054_small'],\n",
    "         ['6089-188582377_tiny'],\n",
    "         ['15708-266043570_medium'],\n",
    "         ['18902-297379468_tiny'],\n",
    "         ['22352-326722819_medium'],\n",
    "         ['25694-352026464_small'],\n",
    "         ['40487-425442645_small'],\n",
    "         ['43448-436041766_tiny'],\n",
    "         ['122213-724743622_tiny'],\n",
    "         ['124277-730508581_medium'],\n",
    "         ['131398-750568980_tiny'],\n",
    "         ['135422-761676527_small'],\n",
    "         ['135578-762107314_medium'],\n",
    "         ['166284-834580727_tiny'],\n",
    "         ['11440-230272102_tiny'],\n",
    "         ['14668-258508783_tiny'],\n",
    "         ['9289-218679635_tiny'],\n",
    "         ['16986-277920487_tiny'],\n",
    "         ['8560-211218186_tiny'],\n",
    "         ['18020-287831424_tiny'],\n",
    "         ['81089-574743647_medium'],\n",
    "        ]\n",
    "\n",
    "lists = pexel_lists + pixabay_lists\n",
    "\n",
    "for ls in lists:\n",
    "    lab = ls[0]\n",
    "    video_path = 'data/time_lapse_video/'+lab+'.mp4'\n",
    "    \n",
    "    # Check if file exists\n",
    "    if os.path.exists('Final_Figs/from_image_analysis/'+lab+'.pdf'):\n",
    "        print('Already exists:', lab)\n",
    "        continue\n",
    "        \n",
    "    else:\n",
    "        print(ls)\n",
    "\n",
    "    \n",
    "    num_bins = 256\n",
    "    EMD1, EMD1_frames, first_frame, max_rds_frame, max_rds_value = process_video_with_metric(video_path, 'EMD', num_bins)\n",
    "    HI1, HI1_frames, first_frame, max_rds_frame, max_rds_value = process_video_with_metric(video_path, '1 - HI', num_bins)\n",
    "    CS1, CS1_frames, first_frame, max_rds_frame, max_rds_value = process_video_with_metric(video_path, 'Chi-Square', num_bins)\n",
    "    KL1, KL1_frames, first_frame, max_rds_frame, max_rds_value = process_video_with_metric(video_path, 'KL', num_bins)\n",
    "    KS1, KS1_frames, first_frame, max_rds_frame, max_rds_value = process_video_with_metric(video_path, 'KS', num_bins)\n",
    "    RPS1, RPS1_frames, first_frame, max_rds_frame, max_rds_value = process_video_with_metric(video_path, 'RPS', num_bins)\n",
    "    RDS1, RDS1_frames, first_frame, max_rds_frame, max_rds_value = process_video_with_metric(video_path, 'RDS', num_bins)\n",
    "\n",
    "    inputs = [[RDS1, RDS1_frames, 'RDS'],\n",
    "              [EMD1, EMD1_frames, 'EMD'],\n",
    "              [HI1, HI1_frames, '1 - HI'],\n",
    "              [CS1, CS1_frames, 'CS'],\n",
    "              [KL1, KL1_frames, 'KL'],\n",
    "              [KS1, KS1_frames, 'KS'],\n",
    "              [RPS1, RPS1_frames, 'RPS']]\n",
    "    make_fig(inputs, lab, first_frame, max_rds_frame, max_rds_value)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1d5f1a42",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}