forked from Siriuscy/MIXLAB_NASA_TICKET
-
Notifications
You must be signed in to change notification settings - Fork 0
/
demo.py
116 lines (100 loc) · 5.77 KB
/
demo.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
from utils_.utils import body_img_path_map, paste, img_pipline, circle, combine_image_to_video, transform_video_to_image
import paddlehub as hub
import matplotlib.pyplot as plt
import cv2
import numpy as np
from PIL import Image
import os
pose_estimation = hub.Module(name="human_pose_estimation_resnet50_mpii")
def pipline(pose_img_path):
result = pose_estimation.keypoint_detection(paths=[pose_img_path])
img_flag = Image.open(body_img_path_map['background']).convert('RGBA')
proportion_shoulder = (1050 - 60) / (result[0]['data']['left_shoulder'][0] - result[0]['data']['right_shoulder'][0])
# read
# body_img = Image.open(body_img_path_map['body']).convert('RGBA')
# body_img = body_img.resize(
# (int(body_img.size[0] / proportion_shoulder), int(body_img.size[1] / proportion_shoulder)))
# helmet_rear_img = Image.open(body_img_path_map['helmet_rear']).convert('RGBA')
# helmet_rear_img = helmet_rear_img.resize(
# (int(helmet_rear_img.size[0] / proportion_shoulder), int(helmet_rear_img.size[1] / proportion_shoulder)))
# helmet_front_img = Image.open(body_img_path_map['helmet_front']).convert('RGBA')
# helmet_front_img = helmet_front_img.resize(
# (int(helmet_front_img.size[0] / proportion_shoulder), int(helmet_front_img.size[1] / proportion_shoulder)))
# bag = Image.open(body_img_path_map['bag']).convert('RGBA')
# bag = bag.resize(
# (int(bag.size[0] / proportion_shoulder), int(bag.size[1] / proportion_shoulder)))
# face_img = circle(body_img_path_map['face'], int(helmet_front_img.size[1]))
# location
# body_x = int(
# (result[0]['data']['left_shoulder'][0] + result[0]['data']['ç'][0]) / 2 - (body_img.size[0] / 2))
# body_y = int(result[0]['data']['left_shoulder'][1] - 550 / 1144 * body_img.size[1])
#
# helmet_x = int((result[0]['data']['left_shoulder'][0] + result[0]['data']['right_shoulder'][0]) / 2 - (
# helmet_front_img.size[0] / 2))
# helmet_y = int(result[0]['data']['left_shoulder'][1] - 1.2 * helmet_front_img.size[1])
#
# face_x = int((result[0]['data']['left_shoulder'][0] + result[0]['data']['right_shoulder'][0]) / 2 - (
# face_img.size[0] / 2))
# face_y = int(result[0]['data']['left_shoulder'][1] - 1.23 * face_img.size[1])
#
# test
# head
img_flag = img_pipline(body_img_path_map['left_elbow'], img_flag, result[0]['data']['left_shoulder'],
result[0]['data']['left_elbow'], key_y_proportion=1 / 4,
img_proportion=proportion_shoulder)
img_flag = img_pipline(body_img_path_map['right_elbow'], img_flag, result[0]['data']['right_shoulder'],
result[0]['data']['right_elbow'], key_y_proportion=1 / 3,
img_proportion=proportion_shoulder)
img_flag = img_pipline(body_img_path_map['body'], img_flag, result[0]['data']['upper_neck'],
result[0]['data']['pelvis'], key_y_proportion=1 / 3,
img_proportion=proportion_shoulder)
img_flag = img_pipline(body_img_path_map['helmet_rear'], img_flag, result[0]['data']['head_top'],
result[0]['data']['upper_neck'], key_y_proportion=1 / 2,
img_proportion=proportion_shoulder)
img_flag = img_pipline(body_img_path_map['face'], img_flag, result[0]['data']['head_top'],
result[0]['data']['upper_neck'], key_y_proportion=12 / 20,
img_proportion=proportion_shoulder)
img_flag = img_pipline(body_img_path_map['helmet_front'], img_flag, result[0]['data']['head_top'],
result[0]['data']['upper_neck'], key_y_proportion=1 / 2,
img_proportion=proportion_shoulder)
# img_flag = paste(bag, img_flag, body_x-13, body_y-30)
# img_flag = paste(body_img, img_flag, body_x, body_y)
# img_flag = paste(helmet_rear_img, img_flag, helmet_x, helmet_y)
# img_flag = paste(face_img, img_flag, face_x, face_y)
# img_flag = paste(helmet_front_img, img_flag, helmet_x, helmet_y)
img_flag = img_pipline(body_img_path_map['left_wrist'], img_flag, result[0]['data']['left_elbow'],
result[0]['data']['left_wrist'], key_y_proportion=150 / 1139,
img_proportion=proportion_shoulder)
img_flag = img_pipline(body_img_path_map['right_wrist'], img_flag, result[0]['data']['right_elbow'],
result[0]['data']['right_wrist'], key_y_proportion=150 / 1139,
img_proportion=proportion_shoulder)
ticket = Image.open(body_img_path_map['ticket']).convert("RGBA")
img_flag = paste(ticket, img_flag, 0, 0)
return img_flag
def analysis_pose(input_frame_path, output_frame_path, is_print=True):
'''
分析图片中的人体姿势， 并转换为皮影姿势，输出结果
'''
file_items = os.listdir(input_frame_path)
file_len = len(file_items)
file_items.sort()
for i in range(file_len):
if is_print:
print(i + 1, '/', file_len, ' ', os.path.join(output_frame_path, str(i) + '.jpg'))
try:
combine_img = pipline(os.path.join(input_frame_path, str(i) + '.jpg'))
combine_img.save(os.path.join(output_frame_path, str(i) + '.png'))
except:
continue
if __name__ == '__main__':
# 第一步：把视频切帧
# fps = transform_video_to_image('./mp4/demo_1.mp4', './mp4_img/')
# 第二步：每一帧都生成一张船票
# analysis_pose('./mp4_img', './out_put', is_print=True)
# 第三步：做成视频:
# combine_image_to_video('./out_put', './out.mp4', 30)
# output one pic
img = pipline('./mp4_img/17.jpg')
print(img.size)
img.save('./one_ticket/one.png')
img.show()