action-recognition-with-skeleton/data_preprocessing.py at master · Mostro-Complexity/action-recognition-with-skeleton · 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
import os
import pickle

import numpy as np
from sklearn.preprocessing import LabelEncoder

ORIGINAL_DATA_PATH = 'data/original'
INPUT_DATA_PATH = 'data/input'

ACTION_FILE_NAMES = {
    '0510160858': 'still',
    '0510161326': 'talking on the phone',
    '0510165136': 'writing on whiteboard',
    '0510161658': 'drinking water',
    '0510171120': 'rinsing mouth with water',
    '0510170707': 'brushing teeth',
    '0510171427': 'wearing contact lenses',
    '0510171507': 'wearing contact lenses',
    '0510162529': 'talking on couch',
    '0510162821': 'relaxing on couch',
    '0510164129': 'cooking (chopping)',
    '0510163840': 'cooking (stirring)',
    '0510163444': 'opening pill container',
    '0510163513': 'opening pill container',
    '0510163542': 'opening pill container',
    '0510164621': 'working on computer',
    '0511121410': 'still',
    '0511121542': 'talking on the phone',
    '0511124850': 'writing on whiteboard',
    '0511121954': 'drinking water',
    '0511130523': 'rinsing mouth with water',
    '0511130138': 'brushing teeth',
    '0511130920': 'wearing contact lenses',
    '0511131018': 'wearing contact lenses',
    '0511122214': 'talking on couch',
    '0511122813': 'relaxing on couch',
    '0511124349': 'cooking (chopping)',
    '0511124101': 'cooking (stirring)',
    '0511123142': 'opening pill container',
    '0511123218': 'opening pill container',
    '0511123238': 'opening pill container',
    '0511123806': 'working on computer',
    '0512172825': 'still',
    '0512171649': 'talking on the phone',
    '0512175502': 'writing on whiteboard',
    '0512173312': 'drinking water',
    '0512164800': 'rinsing mouth with water',
    '0512164529': 'brushing teeth',
    '0512165243': 'wearing contact lenses',
    '0512165327': 'wearing contact lenses',
    '0512174513': 'talking on couch',
    '0512174643': 'relaxing on couch',
    '0512171207': 'cooking (chopping)',
    '0512171444': 'cooking (stirring)',
    '0512173520': 'opening pill container',
    '0512173548': 'opening pill container',
    '0512173623': 'opening pill container',
    '0512170134': 'working on computer',
    '0512150222': 'still',
    '0512150451': 'talking on the phone',
    '0512154505': 'writing on whiteboard',
    '0512150912': 'drinking water',
    '0512155606': 'rinsing mouth with water',
    '0512155226': 'brushing teeth',
    '0512160143': 'wearing contact lenses',
    '0512160254': 'wearing contact lenses',
    '0512151230': 'talking on couch',
    '0512151444': 'relaxing on couch',
    '0512152943': 'cooking (chopping)',
    '0512152416': 'cooking (stirring)',
    '0512151857': 'opening pill container',
    '0512151934': 'opening pill container',
    '0512152013': 'opening pill container',
    '0512153758': 'working on computer'
}


def parse_skeleton_text(line):
    # Parse line by comma
    fields = line.split(',')
    # assert len(fields) == 172, 'Actual length is: ' + str(len(fields))

    frame_num = fields[0]
    skeleton_coords = []

    offset = 1
    for joint_id in range(1, 16):  # 1, 2,...,11
        if joint_id <= 11:  # 1, 2,...,11
            offset += 10  # skip orientation and conf

        x = float(fields[offset])
        offset += 1
        y = float(fields[offset])
        offset += 1
        z = float(fields[offset])
        offset += 1
        conf = float(fields[offset])
        offset += 1

        # pixel_x, pixel_y = pixel_from_coords(x, y, z)
        skeleton_coords.append([x, y, z])

    skeleton_coords = np.array(skeleton_coords)
    return frame_num, skeleton_coords  # ",".join((str(v) for v in output))


if __name__ == "__main__":
    filenames = list(ACTION_FILE_NAMES.keys())
    sample, text_label = [], []

    for fn in filenames:
        with open(os.path.join(ORIGINAL_DATA_PATH, fn + '.txt'), 'r') as f:
            content = [line.rstrip() for line in f.readlines()]
            content.pop()

            all_coords = np.array([parse_skeleton_text(
                line)[1] for line in content])

            video_id = os.path.splitext(fn)[0]

            sample.append(all_coords)
            text_label.append(ACTION_FILE_NAMES[fn])

    label = LabelEncoder().fit_transform(text_label)

    label_dict = {}  # 文字标签到数字标签的对应
    for i in range(len(text_label)):
        label_dict[label[i]] = text_label[i]

    dataset = {
        'label': label,
        'sample': sample
    }

    pickle.dump(dataset, open(os.path.join(
        INPUT_DATA_PATH, 'input.pkl'), 'wb'))
    pickle.dump(label_dict, open(os.path.join(
        INPUT_DATA_PATH, 'label_dict.pkl'), 'wb'))