HumanSkinSegmentation/prepAMI.py at master · gbirman/HumanSkinSegmentation · 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
145
146
147
148
149
150
151
152
153
154
155
156
157

import cv2
import numpy as np
from pathlib import Path
import matplotlib.pyplot as plt
import pickle
from matchVidAMI import getImgPath, getVidPath, img_idx_to_vid_idx, img_folder, vid_folder
import logging

# log and print info simultaneously
def log(msg):
    print(msg)
    logger.error(msg)

# load in dictionaries
load_folder = Path("/Users/gabrielbirman/COS429/SkinSegmentation/Matching/AMI")
img_idx_to_frame_idx_train = np.load(load_folder / 'img_idx_to_frame_idx_train.pickle', allow_pickle=True)
img_idx_to_frame_idx_test = np.load(load_folder / 'img_idx_to_frame_idx_test.pickle', allow_pickle=True)

min_step, max_step = 25 // 4, 25 * 5
step_num = 3
num_inp = 10
save_folder = Path("/Users/gabrielbirman/COS429/SkinSegmentation/Data/VidSequences/AMI")
inp_list, gt_list = [], []
train_idx, test_idx = [], []
delta = []

logging.basicConfig(filename=f"/Users/gabrielbirman/COS429/SkinSegmentation/Data/VidSequences/AMI/{step_num}.txt", filemode='w', format="%(asctime)s;%(message)s", level=logging.ERROR)
logger = logging.getLogger()

# training
for img_idx, vid_idx in img_idx_to_vid_idx.items():

    # get video
    vid_path = getVidPath(*vid_idx, vid_folder)
    vid = cv2.VideoCapture(str(vid_path))

    flag = True
    num_attempts = 0
    num_good = 0
    while flag:
        try:
            step_size = np.random.randint(min_step, max_step)

            # train
            time_series = []

            startFrame = img_idx_to_frame_idx_train[img_idx]

            # frames before center
            for i in range(1, step_num//2 + 1).__reversed__():
                vid.set(1, startFrame - i * step_size)
                _, frame = vid.read(startFrame)
                _, buf = cv2.imencode(".png", frame)
                time_series.append(buf)

            # center frame
            vid.set(1, startFrame)
            _, frame = vid.read()
            _, buf = cv2.imencode(".png", frame)
            time_series.append(buf)

            # frames after center
            for i in range(1, step_num//2 + 1):
                vid.set(1, startFrame + i * step_size)
                _, frame = vid.read(startFrame)
                _, buf = cv2.imencode(".png", frame)
                time_series.append(buf)

            inp_list.append(time_series)

            # groundtruth
            img_path = getImgPath(img_idx, 'train', 'ann', img_folder)
            img = cv2.imread(str(img_path))
            img = np.all(img == [0, 0, 255], axis=2)
            gt_list.append(img)

            log(f"Train index: {img_idx}, step size = {step_size}")
            train_idx.append(len(inp_list)-1)
            delta.append(step_size)

            num_good += 1
            if num_good >= num_inp:
                flag = False
        except:
            num_attempts += 1
            if num_attempts >= 100:
                log(f"Invalid train index: {img_idx}, Start Frame = {startFrame}")
                flag = False

# testing
for img_idx, vid_idx in img_idx_to_vid_idx.items():

    # get video
    vid_path = getVidPath(*vid_idx, vid_folder)
    vid = cv2.VideoCapture(str(vid_path))

    flag = True
    num_attempts = 0
    num_good = 0
    while flag:
        try:
            step_size = np.random.randint(min_step, max_step)

            # train
            time_series = []

            startFrame = img_idx_to_frame_idx_test[img_idx]

            # frames before center
            for i in range(1, step_num//2 + 1).__reversed__():
                vid.set(1, startFrame - i * step_size)
                _, frame = vid.read(startFrame)
                _, buf = cv2.imencode(".png", frame)
                time_series.append(buf)

            # center frame
            vid.set(1, startFrame)
            _, frame = vid.read()
            _, buf = cv2.imencode(".png", frame)
            time_series.append(buf)

            # frames after center
            for i in range(1, step_num//2 + 1):
                vid.set(1, startFrame + i * step_size)
                _, frame = vid.read(startFrame)
                _, buf = cv2.imencode(".png", frame)
                time_series.append(buf)

            inp_list.append(time_series)

            # groundtruth
            img_path = getImgPath(img_idx, 'test', 'ann', img_folder)
            img = cv2.imread(str(img_path))
            img = np.all(img == [0, 0, 255], axis=2)
            gt_list.append(img)

            log(f"Test index: {img_idx}, step size = {step_size}")
            test_idx.append(len(inp_list)-1)
            delta.append(step_size)

            num_good += 1
            if num_good >= num_inp:
                flag = False
        except:
            num_attempts += 1
            if num_attempts >= 100:
                log(f"Invalid test index: {img_idx}, Start Frame = {startFrame}")
                flag = False

data_file = save_folder / f"AMI_Data_{step_num}.npz"
np.savez_compressed(data_file, inp=inp_list, gt=gt_list)
np.savez(save_folder / f"AMI_Data_{step_num}_indices.npz", train=train_idx, test=test_idx)
np.savez(save_folder / f"AMI_Data_{step_num}_delta.npz", delta=delta)

if __name__ == "__main__":
    pass