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
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
| """
安全带误用检测系统架构
基于CUPRA Raval Euro NCAP 2026评测
检测能力:
1. 安全带扣合状态 (SBM-01)
2. 安全带后方佩戴 (SBM-02) ← Raval的核心优势
3. 后排安全带提醒全覆盖
技术方案推测:
- 座椅压力传感器 → 乘员存在
- 安全带张力传感器 → 扣合状态
- B柱/A柱摄像头 → 安全带路径检测
- 安全带 buckle传感器 → 扣合确认
"""
import numpy as np
from typing import Tuple, List
from dataclasses import dataclass
@dataclass
class BeltStatus:
"""安全带状态"""
is_buckled: bool
is_worn_correctly: bool
is_occupied: bool
misuse_type: str
confidence: float
class SeatbeltMisuseDetector:
"""
安全带误用检测器
核心检测逻辑:
1. 座椅压力传感器确认乘员存在
2. Buckle传感器确认安全带扣合
3. 视觉检测安全带路径
4. 路径异常 → 误用告警
关键场景 - 安全带后方佩戴 (SBM-02):
- 正确: 安全带跨过胸前，从左肩到右腰
- 误用: 安全带从身后绕过，未跨胸前
视觉检测关键: 安全带在胸前区域是否可见
"""
def __init__(self):
self.confirmation_frames = 3
self.misuse_count = 0
self.normal_count = 0
self.correct_path_points = [
"shoulder_anchor",
"chest_crossing",
"waist_anchor",
"buckle",
]
def detect_misuse(self,
pressure_sensor: float,
buckle_status: bool,
belt_path_image: np.ndarray,
chest_region_bbox: Tuple[int,int,int,int]) -> BeltStatus:
"""
安全带误用检测主函数
Args:
pressure_sensor: 座椅压力传感器读数 (kg)
buckle_status: 安全带扣合状态
belt_path_image: B柱摄像头图像
chest_region_bbox: 胸前区域边界框 (x1,y1,x2,y2)
Returns:
BeltStatus: 安全带状态
"""
is_occupied = pressure_sensor > 15.0
if not is_occupied:
return BeltStatus(False, True, False, "empty", 1.0)
if not buckle_status:
return BeltStatus(False, False, True, "unbuckled", 1.0)
x1, y1, x2, y2 = chest_region_bbox
chest_region = belt_path_image[y1:y2, x1:x2]
belt_in_chest = self._detect_belt_in_region(chest_region)
if belt_in_chest:
self.normal_count += 1
self.misuse_count = 0
misuse_type = "correct"
confidence = 0.95
else:
self.misuse_count += 1
self.normal_count = 0
if self.misuse_count >= self.confirmation_frames:
misuse_type = "behind_back"
confidence = 0.85
else:
misuse_type = "checking"
confidence = 0.5
return BeltStatus(
is_buckled=True,
is_worn_correctly=(misuse_type == "correct"),
is_occupied=True,
misuse_type=misuse_type,
confidence=confidence
)
def _detect_belt_in_region(self, region: np.ndarray) -> bool:
"""
检测区域内是否有安全带
安全带特征:
- 暗色带状区域 (RGB值低)
- 斜向条带 (约30-60度)
- 宽度约15-30像素
"""
if region.size == 0:
return False
gray = np.mean(region, axis=2) if len(region.shape) == 3 else region
dark_mask = gray < 80
h, w = gray.shape
dark_ratio = np.sum(dark_mask) / (h * w)
if dark_ratio > 0.15 and dark_ratio < 0.6:
row_sums = np.sum(dark_mask, axis=1)
col_sums = np.sum(dark_mask, axis=0)
if np.std(row_sums) / (np.mean(row_sums) + 1e-6) > 0.5:
return True
return False
if __name__ == "__main__":
detector = SeatbeltMisuseDetector()
img_correct = np.random.randint(0, 255, (100, 100, 3), dtype=np.uint8)
img_correct[40:55, 20:80, :] = 30
img_misuse = np.random.randint(100, 255, (100, 100, 3), dtype=np.uint8)
print("=== 正确佩戴 ===")
for i in range(5):
result = detector.detect_misuse(
pressure_sensor=70.0,
buckle_status=True,
belt_path_image=img_correct,
chest_region_bbox=(0, 0, 100, 100)
)
print(f"帧{i}: {result.misuse_type}, conf={result.confidence:.2f}")
print("\n=== 后方佩戴 ===")
detector2 = SeatbeltMisuseDetector()
for i in range(5):
result = detector2.detect_misuse(
pressure_sensor=70.0,
buckle_status=True,
belt_path_image=img_misuse,
chest_region_bbox=(0, 0, 100, 100)
)
print(f"帧{i}: {result.misuse_type}, conf={result.confidence:.2f}")
|
