Object Tracking on Image

问题

I'm using TF API for object detection to detect an object in my video stream. The issue is that even with mobile model, detecting object at every frame slows the video speed. So I was wondering if I can detect the object once and keep tracking it rest of the period. But unfortunately I could not figure out how to do it. I tried opencv tracking APIs, but having issue in tracking the same object over the entire time period. If someone can guide me, it will be of great help.

Thank you!

My working code:

# What model to download.
MODEL_NAME = 'ssd_mobilenet_v1_coco_11_06_2017'
MODEL_FILE = MODEL_NAME + '.tar.gz'
DOWNLOAD_BASE = 'http://download.tensorflow.org/models/object_detection/'

# Path to frozen detection graph. This is the actual model that is used for the object detection.
PATH_TO_CKPT = MODEL_NAME + '/frozen_inference_graph.pb'

# List of the strings that is used to add correct label for each box.
PATH_TO_LABELS = os.path.join('data', 'mscoco_label_map.pbtxt')

NUM_CLASSES = 90

detection_graph = tf.Graph()
with detection_graph.as_default():
  od_graph_def = tf.GraphDef()
  with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:
    serialized_graph = fid.read()
    od_graph_def.ParseFromString(serialized_graph)
    tf.import_graph_def(od_graph_def, name='')

label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)

sess = tf.Session(graph = detection_graph)
cap = cv2.VideoCapture("someVide.mp4")
while(cap.isOpened()):
     ret, frame = cap.read()
     if ret == True:
          frame_expanded = np.expand_dims(frame, axis=0)
        image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
        [boxes, scores, classes, num_detections] = [detection_graph.get_tensor_by_name('detection_boxes:0'), \
        detection_graph.get_tensor_by_name('detection_scores:0'), detection_graph.get_tensor_by_name('detection_classes:0'), \
        detection_graph.get_tensor_by_name('num_detections:0')]
        (boxes, scores, classes, num_detections) = sess.run([boxes, scores, classes, num_detections], \
            feed_dict = {image_tensor: frame_expanded})

        b = vis_util.visualize_boxes_and_labels_on_image_array(
            frame,
            np.squeeze(boxes),
            np.squeeze(classes).astype(np.int32),
            np.squeeze(scores),
            category_index,
            use_normalized_coordinates=True,
            line_thickness=8)
   else:
      break

来源：https://stackoverflow.com/questions/45584199/object-tracking-on-image

标签

python

OpenCV

tensorflow

tracking

object-detection