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
| """
OOP检测算法: 基于关键点的乘员姿态估计
输入: OMS摄像头帧 (RGB, 全局快门)
输出: 乘员姿态等级 + 安全气囊控制信号
关键点检测: 头部中心, 左右肩, 左右髋, 左右膝
姿态计算: 躯干角度, 侧倾角度, 前倾距离, submarining风险
"""
import numpy as np
from typing import Tuple
class OOPDetector:
"""乘员异常姿态检测器"""
def __init__(self):
self.KEYPOINTS = [
"head_center", "left_shoulder", "right_shoulder",
"left_hip", "right_hip", "left_knee", "right_knee"
]
self.FORWARD_TILT_THRESHOLD = 20.0
self.BACKWARD_TILT_THRESHOLD = -15.0
self.LATERAL_TILT_THRESHOLD = 15.0
self.SUBMARINING_HIP_OFFSET = 50
def detect_pose(self, keypoints: np.ndarray) -> dict:
"""
从关键点检测乘员姿态
Args:
keypoints: shape=(7, 2), 7个关键点的(x,y)坐标
[head, L_shoulder, R_shoulder, L_hip, R_hip, L_knee, R_knee]
Returns:
pose: {
'torso_angle': 躯干前倾/后仰角度(正=前倾),
'lateral_angle': 侧倾角度(正=右倾),
'forward_offset': 前倾距离(mm),
'submarining_risk': 0-1,
'oop_level': 0=正常, 1=轻度, 2=中度, 3=重度,
'airbag_mode': 'low_power'|'high_power'|'standard'|'disable'
}
"""
head, l_sh, r_sh, l_hip, r_hip, l_knee, r_knee = keypoints
shoulder_mid = (l_sh + r_sh) / 2
hip_mid = (l_hip + r_hip) / 2
torso_vec = shoulder_mid - hip_mid
torso_angle = np.degrees(np.arctan2(torso_vec[0], -torso_vec[1]))
shoulder_vec = r_sh - l_sh
lateral_angle = np.degrees(np.arctan2(shoulder_vec[1], shoulder_vec[0]))
forward_offset = abs(shoulder_mid[0] - hip_mid[0])
knee_mid = (l_knee + r_knee) / 2
hip_knee_vec = knee_mid - hip_mid
hip_forward = hip_knee_vec[1] < -20
submarining_risk = 0.0
if hip_forward and torso_angle > 10:
submarining_risk = 0.7
elif hip_forward:
submarining_risk = 0.4
elif torso_angle > self.FORWARD_TILT_THRESHOLD:
submarining_risk = 0.3
oop_level = 0
airbag_mode = "standard"
if torso_angle > self.FORWARD_TILT_THRESHOLD:
oop_level = 2
airbag_mode = "low_power"
elif torso_angle < self.BACKWARD_TILT_THRESHOLD:
oop_level = 1
airbag_mode = "high_power"
if abs(lateral_angle) > self.LATERAL_TILT_THRESHOLD:
oop_level = max(oop_level, 2)
airbag_mode = "low_power"
if submarining_risk > 0.5:
oop_level = 3
airbag_mode = "disable"
return {
'torso_angle': round(float(torso_angle), 1),
'lateral_angle': round(float(lateral_angle), 1),
'forward_offset': round(float(forward_offset), 1),
'submarining_risk': round(float(submarining_risk), 2),
'oop_level': oop_level,
'airbag_mode': airbag_mode
}
if __name__ == "__main__":
detector = OOPDetector()
normal_pose = np.array([
[320, 100],
[280, 200],
[360, 200],
[290, 350],
[350, 350],
[285, 500],
[355, 500],
], dtype=np.float32)
result = detector.detect_pose(normal_pose)
print(f"正常坐姿: OOP={result['oop_level']}, 模式={result['airbag_mode']}, 躯干角度={result['torso_angle']}°")
forward_pose = np.array([
[340, 120],
[310, 210],
[370, 210],
[290, 350],
[350, 350],
[285, 500],
[355, 500],
], dtype=np.float32)
result2 = detector.detect_pose(forward_pose)
print(f"前倾坐姿: OOP={result2['oop_level']}, 模式={result2['airbag_mode']}, 躯干角度={result2['torso_angle']}°")
submarining_pose = np.array([
[330, 130],
[295, 230],
[365, 230],
[310, 340],
[370, 340],
[280, 480],
[360, 480],
], dtype=np.float32)
result3 = detector.detect_pose(submarining_pose)
print(f"Submarining风险: OOP={result3['oop_level']}, 模式={result3['airbag_mode']}, 风险={result3['submarining_risk']}")
|
