-
Notifications
You must be signed in to change notification settings - Fork 0
/
parse_xml.py
169 lines (139 loc) · 4.93 KB
/
parse_xml.py
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
158
159
160
161
162
163
164
165
166
167
168
169
import pandas as pd
import cv2
import os
import matplotlib.pyplot as plt
import csv
from pprint import pprint
vid_home = '/home/austin/datasets/MARE/raw'
frame_home = '/home/austin/datasets/MARE/frames'
lid_to_offset = {'517_1160': -1851045, '520_1140': -2007187, '524_1170': -1615301, '532_620': -46680}
def extract_frames(lid, frames):
video_path = os.path.join(vid_home, str(lid) + '.VOB')
dest = os.path.join(frame_home, str(lid))
if not os.path.isdir(dest):
os.mkdir(dest)
else:
print('warning: {} already exists'.format(dest))
# Opens the Video file
cap= cv2.VideoCapture(video_path)
i=0
while(cap.isOpened()):
ret, frame = cap.read()
if ret == False:
break
if i in frames:
print('{}_{}.png'.format(lid, i))
cv2.imwrite( os.path.join(dest, '{}_{}.png'.format(lid, i)), frame)
i+=1
cap.release()
cv2.destroyAllWindows()
def get_frame_num(lid, tc):
# this will fail if we go to next day or something
FPS = 30
offset = lid_to_offset[lid]
hour = tc.hour
minute = tc.minute
second = tc.second
seconds = 3600*hour + 60*minute + second
return (seconds*FPS + offset)
def plot_stats(dfs):
for df_key in ['Fish', 'Inverts']:
if df_key == 'pass':
continue
print(df_key)
per_class_counts = {}
df = dfs[df_key]
spec_set = set()
for i in range(len(df['CommonName'])):
cn = df['CommonName'][i]
if cn in spec_set:
per_class_counts[cn] += 1
else:
per_class_counts[cn] = 1
spec_set.add(df['CommonName'][i])
print(len(spec_set))
pprint(per_class_counts)
xs = []
ys = []
for k,v in per_class_counts.items():
xs.append(k[:9])
ys.append(v)
# plt.hist(ys)
# plt.scatter(xs, ys)
plt.bar(xs, ys)
plt.show()
def organize_dfs(dfs):
lineID_2_f2IDs = {}
# for df_key in dfs.keys():
IDs = set()
for df_key in ['Fish', 'Inverts']:
# are IDs unique across sheets..? IDs makes sure they are
if df_key == 'Habitat':
pass
df = dfs[df_key]
for i in range(len(df)):
lineID = df['LineID'][i]
frame_num = get_frame_num(lineID, df['TC'][i])
if lineID not in lineID_2_f2IDs.keys():
lineID_2_f2IDs[lineID] = {}
frame_to_IDs = lineID_2_f2IDs[lineID]
assert(df['ID'][i] not in IDs)
IDs.add(df['ID'][i])
if frame_num not in frame_to_IDs.keys():
frame_to_IDs[frame_num] = [ (df_key, i, df['ID'][i]) ]
else:
frame_to_IDs[frame_num].append( (df_key, i, df['ID'][i]) )
return lineID_2_f2IDs
def generate_rows(NCOLS, dfs, lineID_2_f2IDs):
header_row = ['filename', 'frame_number', 'survey_date', 'line_id', 'species_1',
'species_1_count', 'species_2', 'species_2_count', 'species_3', 'species_3_count']
rows = []
rows.append(header_row)
for lineID in lineID_2_f2IDs.keys():
frame_to_IDs = lineID_2_f2IDs[lineID]
for frame, IDs in frame_to_IDs.items():
row = []
species = [None]*NCOLS
counts = [None]*NCOLS
fname = '{}_{}.png'.format(lineID, frame)
for i, id_tup in enumerate(IDs):
df_key, df_ind, anno_id = id_tup
df = dfs[df_key]
survey_date = df['SurveyDate'][df_ind]
species[i] = df['CommonName'][df_ind]
counts[i] = df['Count'][df_ind]
row.append(fname)
row.append(frame)
row.append(survey_date)
row.append(lineID)
for i in range(NCOLS):
row.append(species[i])
row.append(counts[i])
rows.append(row)
return rows
if __name__ == "__main__":
xlsx_name = '../raw/Fish_Invert_Habitat_Data.xlsx'
dfs = pd.read_excel(xlsx_name, sheet_name=None)
# plot_stats(dfs)
# map each lineID to a dict that maps frame to (idx, ID) of whats n the frame
lineID_2_f2IDs = organize_dfs(dfs)
# for each video, list frames, and extract them
for lid in lineID_2_f2IDs.keys():
frame_to_IDs = lineID_2_f2IDs[lid]
frames = list(frame_to_IDs.keys())
# extract_frames(lid, frames)
# count number of annos in each frame, and find max
frame_2_count = {}
counts = []
for lineID in lineID_2_f2IDs.keys():
f2IDs = lineID_2_f2IDs[lineID]
for frame in f2IDs.keys():
frame_2_count[frame] = len(f2IDs[frame])
counts.append(len(f2IDs[frame]))
ncols = max(counts)
rows = generate_rows(ncols, dfs, lineID_2_f2IDs)
print(len(rows))
# write rows to our csv file
with open('MARE.csv', 'w') as f:
writer = csv.writer(f)
writer.writerows(rows)