monitoring_vehicle_pollution/OutputImagePainter.py at master · DoMondo/monitoring_vehicle_pollution · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
import cv2
import numpy as np
from matplotlib import pyplot as plt

import VehicleDetector
from CameraCalibration import CameraCalibration

SCALE_FACTOR = 1
IMAGE_W = 1920 // SCALE_FACTOR
IMAGE_H = 1080 // SCALE_FACTOR


def paint_image_detector_tracker(frame, detections, trackings, camera_calibration: CameraCalibration):
    ratio_w = IMAGE_W / frame.shape[1]
    ratio_h = IMAGE_H / frame.shape[0]
    frame = cv2.resize(frame.copy(), (IMAGE_W, IMAGE_H), cv2.INTER_AREA)
    cmap = plt.get_cmap('tab20b')
    colors = [cmap(i)[:3] for i in np.linspace(0, 1, 20)]
    for idx, detection in enumerate(detections[-1]):
        color = colors[int(idx) % len(colors)]
        color = [i * 255 for i in color]
        bbox = detection[:4]
        name = VehicleDetector.class_names[int(detection[5])]
        p0y = int(bbox[1] * IMAGE_H)
        p0x = int(bbox[0] * IMAGE_W)
        p1y = int(bbox[3] * IMAGE_H) + p0y
        p1x = int(bbox[2] * IMAGE_W) + p0x
        cv2.rectangle(frame, (p0x, p0y), (p1x, p1y), color, 6 // SCALE_FACTOR)
        # cv2.putText(frame, str(name) + "-" + str(idx), (int(p0x), int(p0y - 10)), 0, 0.5, (255, 255, 255), 2)
    # Draw detection center and code time with color
    cmap = plt.get_cmap('jet')
    colors = [cmap(i)[:3] for i in np.linspace(0, 1, len(detections))]
    for idx, frame_detections in enumerate(detections):
        color = np.array(colors[idx])
        color *= 255
        for detection in frame_detections:
            screen_pos = camera_calibration.get_centroid_position(detection)
            screen_pos = np.uint16(screen_pos * (ratio_w, ratio_h))
            cv2.circle(frame, screen_pos, 9 // SCALE_FACTOR, color, cv2.FILLED, cv2.LINE_AA)
    # Draw trackings
    cmap = plt.get_cmap('tab20b')
    colors = [cmap(i)[:3] for i in np.linspace(0, 1, 20)]
    for idx, track in enumerate(trackings):
        color = colors[int(idx) % len(colors)]
        color = [i * 255 for i in color]
        for pos_idx in range(len(track) - 1):
            p0 = track[pos_idx][:4]
            p1 = track[pos_idx + 1][:4]
            p0 = camera_calibration.get_centroid_position(p0)
            p1 = camera_calibration.get_centroid_position(p1)
            p0 = np.uint16(p0 * (ratio_w, ratio_h))
            p1 = np.uint16(p1 * (ratio_w, ratio_h))
            # if pos_idx == 0 or pos_idx == len(track) - 1:
            #     cv2.line(frame, p0, p1, [255, 255, 255], 9 // SCALE_FACTOR, lineType=cv2.LINE_AA)
            # else:
            cv2.line(frame, p0, p1, color, 9 // SCALE_FACTOR, lineType=cv2.LINE_AA)

    # Draw first and last points of each tracking
    for idx, track in enumerate(trackings):
        p0 = track[0][:4]
        p1 = track[-1][:4]
        p0 = camera_calibration.get_centroid_position(p0)
        p1 = camera_calibration.get_centroid_position(p1)
        p0 = np.uint16(p0 * (ratio_w, ratio_h))
        p1 = np.uint16(p1 * (ratio_w, ratio_h))
        cv2.circle(frame, p0, 9 // SCALE_FACTOR, [255, 255, 255], 3, lineType=cv2.LINE_AA)
        cv2.circle(frame, p1, 9 // SCALE_FACTOR, [255, 255, 255], 3, lineType=cv2.LINE_AA)

    # Draw goal line
    p0, p1 = camera_calibration.get_goal_line()
    p0[0] *= ratio_w
    p0[1] *= ratio_h
    p1[0] *= ratio_w
    p1[1] *= ratio_h
    p0 = np.uint16(p0)
    p1 = np.uint16(p1)
    cv2.line(frame, p0, p1, (255, 0, 0), 2, lineType=cv2.LINE_AA)
    frame = cv2.cvtColor(frame, cv2.COLOR_RGB2BGR)
    frame = np.uint8(frame)
    return frame


def paint_detector_tracker_top_view(input_image, detections, trackings, camera_calibration: CameraCalibration):
    # Generate top view
    frame = camera_calibration.generate_top_view(input_image, IMAGE_H, IMAGE_W)
    ratio_w = frame.shape[0] / input_image.shape[1]
    ratio_h = frame.shape[1] / input_image.shape[0]
    cmap = plt.get_cmap('jet')
    colors = [cmap(i)[:3] for i in np.linspace(0, 1, len(detections))]
    # Draw Goal Line
    p0, p1 = camera_calibration.get_goal_line()
    p0x, p0y = camera_calibration.get_xyz_top_view(p0[0], p0[1])
    p1x, p1y = camera_calibration.get_xyz_top_view(p1[0], p1[1])
    p0y = np.uint16(p0y)
    p1y = np.uint16(p1y)
    p0x = np.uint16(p0x)
    p1x = np.uint16(p1x)
    cv2.line(frame, (p0x, p0y), (p1x, p1y), (255, 0, 0), 3 // SCALE_FACTOR, lineType=cv2.LINE_AA)
    # Draw detection center and code time with color
    for idx, frame_detections in enumerate(detections):
        color = np.array(colors[idx])
        color *= 255
        for detection in frame_detections:
            screen_pos = camera_calibration.get_centroid_position(detection)
            screen_pos = np.uint16(screen_pos)
            xyz = camera_calibration.get_xyz_top_view(screen_pos[0], screen_pos[1])
            x = np.int16(xyz[0])
            y = np.int16(xyz[1])
            cv2.circle(frame, (x, y), 9 // SCALE_FACTOR, color, cv2.FILLED, cv2.LINE_AA)
    frame = cv2.cvtColor(frame, cv2.COLOR_RGB2BGR)

    for idx, track in enumerate(trackings):
        color = colors[int(idx) % len(colors)]
        color = [i * 255 for i in color]
        for pos_idx in range(len(track) - 1):
            p0 = track[pos_idx][:4]
            p1 = track[pos_idx + 1][:4]
            p0 = camera_calibration.get_centroid_position(p0)
            p1 = camera_calibration.get_centroid_position(p1)
            p0 = np.uint16(p0 * (ratio_w, ratio_h))
            p1 = np.uint16(p1 * (ratio_w, ratio_h))

            xyz = camera_calibration.get_xyz_top_view(p0[0], p0[1])
            p0[0] = np.int16(xyz[0])
            p0[1] = np.int16(xyz[1])
            xyz = camera_calibration.get_xyz_top_view(p1[0], p1[1])
            p1[0] = np.int16(xyz[0])
            p1[1] = np.int16(xyz[1])
            # if pos_idx == 0 or pos_idx == len(track) - 1:
            #     cv2.line(frame, p0, p1, [255, 255, 255], 9 // SCALE_FACTOR, lineType=cv2.LINE_AA)
            # else:
            cv2.line(frame, p0, p1, color, 9 // SCALE_FACTOR, lineType=cv2.LINE_AA)

    # Draw first and last points of each tracking
    for idx, track in enumerate(trackings):
        p0 = track[0][:4]
        p1 = track[-1][:4]
        p0 = camera_calibration.get_centroid_position(p0)
        p1 = camera_calibration.get_centroid_position(p1)
        p0 = np.uint16(p0 * (ratio_w, ratio_h))
        p1 = np.uint16(p1 * (ratio_w, ratio_h))
        cv2.circle(frame, p0, 9 // SCALE_FACTOR, [255, 255, 255], 3, lineType=cv2.LINE_AA)
        cv2.circle(frame, p1, 9 // SCALE_FACTOR, [255, 255, 255], 3, lineType=cv2.LINE_AA)
    return frame