detect_object.py

import os
import numpy as np
from utils import label_map_util
from utils.webcam import draw_boxes_and_labels


CWD_PATH = os.getcwd()
PATH_TO_LABELS = os.path.join(CWD_PATH, 'detection', 'data', 'mscoco_label_map.pbtxt')

NUM_CLASSES = 90
# label map
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)

# pass in image_np, returns 
def detect_objects(image_np, sess, detection_graph):

    image_np_expanded = np.expand_dims(image_np, axis=0)
    image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')

    boxes = detection_graph.get_tensor_by_name('detection_boxes:0')

    scores = detection_graph.get_tensor_by_name('detection_scores:0')
    classes = detection_graph.get_tensor_by_name('detection_classes:0')
    num_detections = detection_graph.get_tensor_by_name('num_detections:0')

    # Do the detection/model prediction here
    (boxes, scores, classes, num_detections) = sess.run(
        [boxes, scores, classes, num_detections],
        feed_dict={image_tensor: image_np_expanded})
    
    rect_points, class_names, class_colors = draw_boxes_and_labels(
        boxes=np.squeeze(boxes),
        classes=np.squeeze(classes).astype(np.int32),
        scores=np.squeeze(scores),
        category_index=category_index,
        min_score_thresh=.5
    )

    return dict(rect_points=rect_points, class_names=class_names, class_colors=class_colors)