{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Object Detection Demo\n", "Welcome to the object detection inference walkthrough! This notebook will walk you step by step through the process of using a pre-trained model to detect objects in an image. Make sure to follow the [installation instructions](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/installation.md) before you start." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Imports" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1.5.0\n" ] } ], "source": [ "import numpy as np\n", "import os\n", "import six.moves.urllib as urllib\n", "import sys\n", "import tarfile\n", "import tensorflow as tf\n", "import zipfile\n", "\n", "from collections import defaultdict\n", "from io import StringIO\n", "from matplotlib import pyplot as plt\n", "from PIL import Image\n", "\n", "# This is needed since the notebook is stored in the object_detection folder.\n", "sys.path.append(\"..\")\n", "from object_detection.utils import ops as utils_ops\n", "\n", "if tf.__version__ < '1.4.0':\n", " raise ImportError('Please upgrade your tensorflow installation to v1.4.* or later!')\n", "print(tf.__version__)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Env setup" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Object detection imports\n", "Here are the imports from the object detection module." ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "from utils import label_map_util\n", "\n", "from utils import visualization_utils as vis_util" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Model preparation " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Variables\n", "\n", "Any model exported using the `export_inference_graph.py` tool can be loaded here simply by changing `PATH_TO_CKPT` to point to a new .pb file. \n", "\n", "By default we use an \"SSD with Mobilenet\" model here. See the [detection model zoo](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md) for a list of other models that can be run out-of-the-box with varying speeds and accuracies." ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [], "source": [ "# Path to frozen detection graph. This is the actual model that is used for the object detection.\n", "PATH_TO_CKPT = '/home/priya/Documents/mask_rcnn/mask_rcnn_inception_v2_coco_2018_01_28' + '/frozen_inference_graph.pb'\n", "\n", "# List of the strings that is used to add correct label for each box.\n", "PATH_TO_LABELS = os.path.join('data', 'mscoco_label_map.pbtxt')\n", "\n", "NUM_CLASSES = 90" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Load a (frozen) Tensorflow model into memory." ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [], "source": [ "detection_graph = tf.Graph()\n", "with detection_graph.as_default():\n", " od_graph_def = tf.GraphDef()\n", " with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:\n", " serialized_graph = fid.read()\n", " od_graph_def.ParseFromString(serialized_graph)\n", " tf.import_graph_def(od_graph_def, name='')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Loading label map\n", "Label maps map indices to category names, so that when our convolution network predicts `5`, we know that this corresponds to `airplane`. Here we use internal utility functions, but anything that returns a dictionary mapping integers to appropriate string labels would be fine" ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [], "source": [ "label_map = label_map_util.load_labelmap(PATH_TO_LABELS)\n", "categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)\n", "category_index = label_map_util.create_category_index(categories)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Helper code" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [], "source": [ "def load_image_into_numpy_array(image):\n", " (im_width, im_height) = image.size\n", " return np.array(image.getdata()).reshape(\n", " (im_height, im_width, 3)).astype(np.uint8)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Detection" ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [], "source": [ "# Import everything needed to edit/save/watch video clips\n", "from moviepy.editor import VideoFileClip\n", "from IPython.display import HTML" ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [], "source": [ "def detect_videos(image_np, sess, detection_graph):\n", " \n", " with detection_graph.as_default():\n", " \n", " ops = tf.get_default_graph().get_operations()\n", " all_tensor_names = {output.name for op in ops for output in op.outputs}\n", " tensor_dict = {}\n", " for key in [\n", " 'num_detections', 'detection_boxes', 'detection_scores',\n", " 'detection_classes', 'detection_masks'\n", " ]:\n", " tensor_name = key + ':0'\n", " if tensor_name in all_tensor_names:\n", " tensor_dict[key] = tf.get_default_graph().get_tensor_by_name(tensor_name)\n", " if 'detection_masks' in tensor_dict:\n", " # The following processing is only for single image\n", " detection_boxes = tf.squeeze(tensor_dict['detection_boxes'], [0])\n", " detection_masks = tf.squeeze(tensor_dict['detection_masks'], [0])\n", " # Reframe is required to translate mask from box coordinates to image coordinates and fit the image size.\n", " real_num_detection = tf.cast(tensor_dict['num_detections'][0], tf.int32)\n", " detection_boxes = tf.slice(detection_boxes, [0, 0], [real_num_detection, -1])\n", " detection_masks = tf.slice(detection_masks, [0, 0, 0], [real_num_detection, -1, -1])\n", " detection_masks_reframed = utils_ops.reframe_box_masks_to_image_masks(\n", " detection_masks, detection_boxes, image_np.shape[0], image_np.shape[1])\n", " detection_masks_reframed = tf.cast(\n", " tf.greater(detection_masks_reframed, 0.5), tf.uint8)\n", " # Follow the convention by adding back the batch dimension\n", " tensor_dict['detection_masks'] = tf.expand_dims(\n", " detection_masks_reframed, 0)\n", " image_tensor = tf.get_default_graph().get_tensor_by_name('image_tensor:0')\n", "\n", " # Run inference\n", " output_dict = sess.run(tensor_dict,\n", " feed_dict={image_tensor: np.expand_dims(image_np, 0)})\n", "\n", " # all outputs are float32 numpy arrays, so convert types as appropriate\n", " output_dict['num_detections'] = int(output_dict['num_detections'][0])\n", " output_dict['detection_classes'] = output_dict[\n", " 'detection_classes'][0].astype(np.uint8)\n", " output_dict['detection_boxes'] = output_dict['detection_boxes'][0]\n", " output_dict['detection_scores'] = output_dict['detection_scores'][0]\n", " if 'detection_masks' in output_dict:\n", " output_dict['detection_masks'] = output_dict['detection_masks'][0]\n", "\n", " vis_util.visualize_boxes_and_labels_on_image_array(\n", " image_np,\n", " output_dict['detection_boxes'],\n", " output_dict['detection_classes'],\n", " output_dict['detection_scores'],\n", " category_index,\n", " instance_masks=output_dict.get('detection_masks'),\n", " use_normalized_coordinates=True,\n", " line_thickness=1)\n", " \n", " return image_np" ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [], "source": [ "def process_image(image): \n", " \n", " global counter\n", " \n", " if counter%1 ==0:\n", " \n", " with detection_graph.as_default():\n", " with tf.Session(graph=detection_graph) as sess:\n", " image_np = detect_videos(image, sess, detection_graph) \n", "\n", " counter +=1 \n", " \n", " return image" ] }, { "cell_type": "code", "execution_count": 20, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[MoviePy] >>>> Building video videos_out/cars_ppl_out.mp4\n", "[MoviePy] Writing video videos_out/cars_ppl_out.mp4\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n", " 0%| | 0/192 [00:00>>> Video ready: videos_out/cars_ppl_out.mp4 \n", "\n", "CPU times: user 1h 22min 33s, sys: 1min 45s, total: 1h 24min 19s\n", "Wall time: 18min 27s\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\u001b[A\n", "\u001b[A" ] } ], "source": [ "filename = 'videos_in/cars_ppl.mp4'\n", "new_loc = 'videos_out/cars_ppl_out.mp4'\n", "\n", "counter = 0\n", "\n", "white_output = new_loc\n", "clip1 = VideoFileClip(filename).subclip(60,68)\n", "white_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!s\n", "%time white_clip.write_videofile(white_output, audio=False)" ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[MoviePy] >>>> Building video videos_out/cars_ppl2_out.mp4\n", "[MoviePy] Writing video videos_out/cars_ppl2_out.mp4\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n", " 0%| | 0/240 [00:00>>> Video ready: videos_out/cars_ppl2_out.mp4 \n", "\n", "CPU times: user 1h 45min 15s, sys: 2min 17s, total: 1h 47min 32s\n", "Wall time: 25min 5s\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\u001b[A\n", "\u001b[A" ] } ], "source": [ "filename = 'videos_in/cars_ppl2.mp4'\n", "new_loc = 'videos_out/cars_ppl2_out.mp4'\n", "\n", "counter = 0\n", "\n", "white_output = new_loc\n", "clip1 = VideoFileClip(filename).subclip(64,72)\n", "white_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!s\n", "%time white_clip.write_videofile(white_output, audio=False)" ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[MoviePy] >>>> Building video videos_out/kid_soccer_out.mp4\n", "[MoviePy] Writing video videos_out/kid_soccer_out.mp4\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n", " 0%| | 0/150 [00:00>>> Video ready: videos_out/kid_soccer_out.mp4 \n", "\n", "CPU times: user 1h 6min 56s, sys: 1min 24s, total: 1h 8min 21s\n", "Wall time: 16min 48s\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\u001b[A\n", "\u001b[A" ] } ], "source": [ "filename = 'videos_in/kid_soccer.mp4'\n", "new_loc = 'videos_out/kid_soccer_out.mp4'\n", "\n", "counter = 0\n", "\n", "white_output = new_loc\n", "clip1 = VideoFileClip(filename).subclip(18,23)\n", "white_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!s\n", "%time white_clip.write_videofile(white_output, audio=False)" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[MoviePy] >>>> Building video videos_out/cat_dog_out.mp4\n", "[MoviePy] Writing video videos_out/cat_dog_out.mp4\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n", " 0%| | 0/151 [00:00>>> Video ready: videos_out/cat_dog_out.mp4 \n", "\n", "CPU times: user 1h 7min 41s, sys: 1min 26s, total: 1h 9min 7s\n", "Wall time: 18min 43s\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\u001b[A\n", "\u001b[A" ] } ], "source": [ "filename = 'videos_in/cat_dog.mp4'\n", "new_loc = 'videos_out/cat_dog_out.mp4'\n", "\n", "counter = 0\n", "\n", "white_output = new_loc\n", "clip1 = VideoFileClip(filename).subclip(120,125)\n", "white_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!s\n", "%time white_clip.write_videofile(white_output, audio=False)" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[MoviePy] >>>> Building video videos_out/banana_out.mp4\n", "[MoviePy] Writing video videos_out/banana_out.mp4\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n", " 0%| | 0/144 [00:00>>> Video ready: videos_out/banana_out.mp4 \n", "\n", "CPU times: user 1h 6min 10s, sys: 1min 28s, total: 1h 7min 39s\n", "Wall time: 18min 42s\n" ] } ], "source": [ "filename = 'videos_in/fruits.mp4'\n", "new_loc = 'videos_out/banana_out.mp4'\n", "\n", "counter = 0\n", "\n", "white_output = new_loc\n", "clip1 = VideoFileClip(filename).subclip(79,85)\n", "white_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!s\n", "%time white_clip.write_videofile(white_output, audio=False)" ] }, { "cell_type": "code", "execution_count": 25, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[MoviePy] >>>> Building video videos_out/apple_picking_out.mp4\n", "[MoviePy] Writing video videos_out/apple_picking_out.mp4\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n", " 0%| | 0/91 [00:00>>> Video ready: videos_out/apple_picking_out.mp4 \n", "\n", "CPU times: user 41min 42s, sys: 54 s, total: 42min 36s\n", "Wall time: 11min 44s\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\u001b[A\n", "\u001b[A" ] } ], "source": [ "filename = 'videos_in/apple_picking.mp4'\n", "new_loc = 'videos_out/apple_picking_out.mp4'\n", "\n", "counter = 0\n", "\n", "white_output = new_loc\n", "clip1 = VideoFileClip(filename).subclip(41,44)\n", "white_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!s\n", "%time white_clip.write_videofile(white_output, audio=False)" ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[MoviePy] >>>> Building video videos_out/horse_out.mp4\n", "[MoviePy] Writing video videos_out/horse_out.mp4\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n", " 0%| | 0/120 [00:00>>> Video ready: videos_out/horse_out.mp4 \n", "\n", "CPU times: user 56min 12s, sys: 1min 10s, total: 57min 22s\n", "Wall time: 16min 5s\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\u001b[A\n", "\u001b[A" ] } ], "source": [ "filename = 'videos_in/horse.mp4'\n", "new_loc = 'videos_out/horse_out.mp4'\n", "\n", "counter = 0\n", "\n", "white_output = new_loc\n", "clip1 = VideoFileClip(filename).subclip(45,49)\n", "white_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!s\n", "%time white_clip.write_videofile(white_output, audio=False)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "colab": { "version": "0.3.2" }, "kernelspec": { "display_name": "Python 2", "language": "python", "name": "python2" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 2 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython2", "version": "2.7.12" } }, "nbformat": 4, "nbformat_minor": 2 }