|
| 1 | +{ |
| 2 | + "cells": [ |
| 3 | + { |
| 4 | + "cell_type": "code", |
| 5 | + "execution_count": 1, |
| 6 | + "metadata": {}, |
| 7 | + "outputs": [], |
| 8 | + "source": [ |
| 9 | + "import numpy as np\n", |
| 10 | + "import cv2 as cv\n", |
| 11 | + "from motrackers import SimpleTracker\n", |
| 12 | + "from motrackers.utils import select_caffemodel, select_videofile" |
| 13 | + ] |
| 14 | + }, |
| 15 | + { |
| 16 | + "cell_type": "code", |
| 17 | + "execution_count": 2, |
| 18 | + "metadata": {}, |
| 19 | + "outputs": [ |
| 20 | + { |
| 21 | + "data": { |
| 22 | + "application/vnd.jupyter.widget-view+json": { |
| 23 | + "model_id": "1d3c3a3a86324074b3461a40bda983ea", |
| 24 | + "version_major": 2, |
| 25 | + "version_minor": 0 |
| 26 | + }, |
| 27 | + "text/plain": [ |
| 28 | + "FileChooser(path='..', filename='', show_hidden='False')" |
| 29 | + ] |
| 30 | + }, |
| 31 | + "metadata": {}, |
| 32 | + "output_type": "display_data" |
| 33 | + }, |
| 34 | + { |
| 35 | + "data": { |
| 36 | + "application/vnd.jupyter.widget-view+json": { |
| 37 | + "model_id": "46b604d3ae4d4853aa0a644f9edfe463", |
| 38 | + "version_major": 2, |
| 39 | + "version_minor": 0 |
| 40 | + }, |
| 41 | + "text/plain": [ |
| 42 | + "FileChooser(path='..', filename='', show_hidden='False')" |
| 43 | + ] |
| 44 | + }, |
| 45 | + "metadata": {}, |
| 46 | + "output_type": "display_data" |
| 47 | + }, |
| 48 | + { |
| 49 | + "data": { |
| 50 | + "application/vnd.jupyter.widget-view+json": { |
| 51 | + "model_id": "9f31f72dfb7446b29e258b97fa8f1a2f", |
| 52 | + "version_major": 2, |
| 53 | + "version_minor": 0 |
| 54 | + }, |
| 55 | + "text/plain": [ |
| 56 | + "FileChooser(path='..', filename='', show_hidden='False')" |
| 57 | + ] |
| 58 | + }, |
| 59 | + "metadata": {}, |
| 60 | + "output_type": "display_data" |
| 61 | + } |
| 62 | + ], |
| 63 | + "source": [ |
| 64 | + "video_file = select_videofile('..')\n", |
| 65 | + "prototxt, weights = select_caffemodel('..')\n", |
| 66 | + "display(video_file, prototxt, weights)" |
| 67 | + ] |
| 68 | + }, |
| 69 | + { |
| 70 | + "cell_type": "code", |
| 71 | + "execution_count": 3, |
| 72 | + "metadata": {}, |
| 73 | + "outputs": [], |
| 74 | + "source": [ |
| 75 | + "video = video_file.selected" |
| 76 | + ] |
| 77 | + }, |
| 78 | + { |
| 79 | + "cell_type": "code", |
| 80 | + "execution_count": 4, |
| 81 | + "metadata": {}, |
| 82 | + "outputs": [], |
| 83 | + "source": [ |
| 84 | + "model = {\"prototxt\": prototxt.selected,\n", |
| 85 | + " \"weights\": weights.selected,\n", |
| 86 | + " \"object_names\": {0: 'background', \n", |
| 87 | + " 1: 'aeroplane', \n", |
| 88 | + " 2: 'bicycle', \n", |
| 89 | + " 3: 'bird',\n", |
| 90 | + " 4: 'boat',\n", |
| 91 | + " 5: 'bottle',\n", |
| 92 | + " 6: 'bus', \n", |
| 93 | + " 7: 'car', \n", |
| 94 | + " 8: 'cat', \n", |
| 95 | + " 9: 'chair',\n", |
| 96 | + " 10: 'cow', \n", |
| 97 | + " 11: 'diningtable', \n", |
| 98 | + " 12: 'dog', \n", |
| 99 | + " 13: 'horse',\n", |
| 100 | + " 14: 'motorbike', \n", |
| 101 | + " 15: 'person', \n", |
| 102 | + " 16: 'pottedplant',\n", |
| 103 | + " 17: 'sheep', \n", |
| 104 | + " 18: 'sofa', \n", |
| 105 | + " 19: 'train',\n", |
| 106 | + " 20: 'tvmonitor'},\n", |
| 107 | + " \"threshold\": 0.2,\n", |
| 108 | + " \"confidence_threshold\": 0.2,\n", |
| 109 | + " \"pixel_std\":1/127.5,\n", |
| 110 | + " \"pixel_mean\": 127.5,\n", |
| 111 | + " \"input_size\": (300, 300)\n", |
| 112 | + " }\n", |
| 113 | + "\n", |
| 114 | + "max_object_lost_count = 5 # maximum number of object losts counted when the object is being tracked\n", |
| 115 | + "\n", |
| 116 | + "np.random.seed(12345)\n", |
| 117 | + "bbox_colors = {key: np.random.randint(0, 255, size=(3,)).tolist() for key in model['object_names'].keys()}" |
| 118 | + ] |
| 119 | + }, |
| 120 | + { |
| 121 | + "cell_type": "code", |
| 122 | + "execution_count": 5, |
| 123 | + "metadata": {}, |
| 124 | + "outputs": [], |
| 125 | + "source": [ |
| 126 | + "cap = cv.VideoCapture(video)\n", |
| 127 | + "net = cv.dnn.readNetFromCaffe(model[\"prototxt\"], model[\"weights\"])\n", |
| 128 | + "tracker = SimpleTracker(max_lost=max_object_lost_count)" |
| 129 | + ] |
| 130 | + }, |
| 131 | + { |
| 132 | + "cell_type": "code", |
| 133 | + "execution_count": 6, |
| 134 | + "metadata": { |
| 135 | + "scrolled": false |
| 136 | + }, |
| 137 | + "outputs": [], |
| 138 | + "source": [ |
| 139 | + "(H, W) = (None, None)\n", |
| 140 | + "writer = None\n", |
| 141 | + "\n", |
| 142 | + "while True:\n", |
| 143 | + " ok, image = cap.read()\n", |
| 144 | + " \n", |
| 145 | + " if not ok:\n", |
| 146 | + " print(\"Cannot read the video feed.\")\n", |
| 147 | + " break\n", |
| 148 | + " \n", |
| 149 | + " if W is None or H is None: \n", |
| 150 | + " (H, W) = image.shape[:2]\n", |
| 151 | + " \n", |
| 152 | + " image_resized = cv.resize(image, model[\"input_size\"])\n", |
| 153 | + "\n", |
| 154 | + " blob = cv.dnn.blobFromImage(image_resized, \n", |
| 155 | + " model[\"pixel_std\"], \n", |
| 156 | + " model[\"input_size\"], \n", |
| 157 | + " (model[\"pixel_mean\"], model[\"pixel_mean\"], model[\"pixel_mean\"]), \n", |
| 158 | + " False)\n", |
| 159 | + "\n", |
| 160 | + " net.setInput(blob)\n", |
| 161 | + " detections = net.forward()\n", |
| 162 | + "\n", |
| 163 | + " rows = image_resized.shape[0]\n", |
| 164 | + " cols = image_resized.shape[1]\n", |
| 165 | + " \n", |
| 166 | + " boxes, confidences, classIDs, detections_bbox = [], [], [], []\n", |
| 167 | + "\n", |
| 168 | + " for i in range(detections.shape[2]):\n", |
| 169 | + " confidence = detections[0, 0, i, 2]\n", |
| 170 | + " if confidence > model['confidence_threshold']:\n", |
| 171 | + " class_id = int(detections[0, 0, i, 1])\n", |
| 172 | + "\n", |
| 173 | + " # object location \n", |
| 174 | + " left = int(detections[0, 0, i, 3] * cols) \n", |
| 175 | + " top = int(detections[0, 0, i, 4] * rows)\n", |
| 176 | + " right = int(detections[0, 0, i, 5] * cols)\n", |
| 177 | + " bottom = int(detections[0, 0, i, 6] * rows)\n", |
| 178 | + " \n", |
| 179 | + " # scaling factor of image\n", |
| 180 | + " height_factor = image.shape[0]/float(model[\"input_size\"][0])\n", |
| 181 | + " width_factor = image.shape[1]/float(model[\"input_size\"][1])\n", |
| 182 | + " \n", |
| 183 | + " # scale object detection bounding box to original image\n", |
| 184 | + " left = int(width_factor * left) \n", |
| 185 | + " top = int(height_factor * top)\n", |
| 186 | + " right = int(width_factor * right)\n", |
| 187 | + " bottom = int(height_factor * bottom)\n", |
| 188 | + " \n", |
| 189 | + " width, height = right - left, bottom-top\n", |
| 190 | + " \n", |
| 191 | + " boxes.append([left, top, width, height])\n", |
| 192 | + " confidences.append(float(confidence))\n", |
| 193 | + " classIDs.append(int(class_id))\n", |
| 194 | + " \n", |
| 195 | + " indices = cv.dnn.NMSBoxes(boxes, confidences, model[\"confidence_threshold\"], model[\"threshold\"])\n", |
| 196 | + " \n", |
| 197 | + " if len(indices)>0:\n", |
| 198 | + " for i in indices.flatten():\n", |
| 199 | + " x, y, w, h = boxes[i][0], boxes[i][1], boxes[i][2], boxes[i][3]\n", |
| 200 | + " \n", |
| 201 | + " detections_bbox.append((x, y, x+w, y+h))\n", |
| 202 | + " \n", |
| 203 | + " clr = [int(c) for c in bbox_colors[classIDs[i]]]\n", |
| 204 | + " cv.rectangle(image, (x, y), (x+w, y+h), clr, 2)\n", |
| 205 | + " \n", |
| 206 | + " label = \"{}:{:.4f}\".format(model[\"object_names\"][classIDs[i]], confidences[i])\n", |
| 207 | + " (label_width, label_height), baseLine = cv.getTextSize(label, cv.FONT_HERSHEY_SIMPLEX, 0.5, 2)\n", |
| 208 | + " y_label = max(y, label_height)\n", |
| 209 | + " cv.rectangle(image, (x, y_label-label_height),\n", |
| 210 | + " (x+label_width, y_label+baseLine), (255, 255, 255), cv.FILLED)\n", |
| 211 | + " cv.putText(image, label, (x, y_label), cv.FONT_HERSHEY_SIMPLEX, 0.5, clr, 2)\n", |
| 212 | + " \n", |
| 213 | + " objects = tracker.update(detections_bbox)\n", |
| 214 | + " \n", |
| 215 | + " for (objectID, centroid) in objects.items():\n", |
| 216 | + " text = \"ID {}\".format(objectID)\n", |
| 217 | + " cv.putText(image, text, (centroid[0] - 10, centroid[1] - 10), cv.FONT_HERSHEY_SIMPLEX,\n", |
| 218 | + " 0.5, (0, 255, 0), 2)\n", |
| 219 | + " cv.circle(image, (centroid[0], centroid[1]), 4, (0, 255, 0), -1)\n", |
| 220 | + " \n", |
| 221 | + " cv.imshow(\"image\", image)\n", |
| 222 | + " \n", |
| 223 | + " if cv.waitKey(1) & 0xFF == ord('q'):\n", |
| 224 | + " break\n", |
| 225 | + " \n", |
| 226 | + " if writer is None:\n", |
| 227 | + " fourcc = cv.VideoWriter_fourcc(*\"MJPG\")\n", |
| 228 | + " writer = cv.VideoWriter(\"output.avi\", fourcc, 30, (W, H), True)\n", |
| 229 | + " writer.write(image)\n", |
| 230 | + "\n", |
| 231 | + "writer.release()\n", |
| 232 | + "cap.release()\n", |
| 233 | + "cv.destroyWindow(\"image\")" |
| 234 | + ] |
| 235 | + }, |
| 236 | + { |
| 237 | + "cell_type": "code", |
| 238 | + "execution_count": null, |
| 239 | + "metadata": {}, |
| 240 | + "outputs": [], |
| 241 | + "source": [] |
| 242 | + } |
| 243 | + ], |
| 244 | + "metadata": { |
| 245 | + "kernelspec": { |
| 246 | + "display_name": "Python 3", |
| 247 | + "language": "python", |
| 248 | + "name": "python3" |
| 249 | + }, |
| 250 | + "language_info": { |
| 251 | + "codemirror_mode": { |
| 252 | + "name": "ipython", |
| 253 | + "version": 3 |
| 254 | + }, |
| 255 | + "file_extension": ".py", |
| 256 | + "mimetype": "text/x-python", |
| 257 | + "name": "python", |
| 258 | + "nbconvert_exporter": "python", |
| 259 | + "pygments_lexer": "ipython3", |
| 260 | + "version": "3.6.9" |
| 261 | + } |
| 262 | + }, |
| 263 | + "nbformat": 4, |
| 264 | + "nbformat_minor": 2 |
| 265 | +} |
0 commit comments