jc/jc-video-recognize
11
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: 新增人员徘徊/静止行为分析功能

Browse Source 
本次提交实现了完整的人员行为分析系统，包括：
1. 新增基于位置和跟踪ID的两种行为检测算法
2. 新增徘徊检测服务与行为处理器模块
3. 前后端集成算法配置界面与告警展示
4. 支持图片和视频流场景下的行为分析
5. 新增算法配置接口与文档说明

具体改动：
- 新增loitering_detection模型目录与算法实现
- 新增AlgorithmConfig组件实现可视化配置
- 扩展图片/视频检测接口支持算法参数传递
- 新增行为告警推送与前端展示页面
- 优化检测服务，集成行为分析逻辑
- 移除冗余日志输出，完善代码注释
This commit is contained in:
wwh
2026-05-19 09:17:09 +08:00
parent 2691761f01
commit 7aa71c5f83
15 changed files with 1937 additions and 76 deletions
Download Patch File Download Diff File Expand all files Collapse all files

103
apps/server/api/detection.py
View File

@@ -2,24 +2,50 @@ import cv2
import numpy as np
import base64
import logging
import json
from typing import Optional
from fastapi import APIRouter, UploadFile, File, Form, Query
from models.schemas import ImageDetectionResult
router = APIRouter()
logger = logging.getLogger(__name__)
@router.post("/detect/image", response_model=ImageDetectionResult)
async def detect_image(
file: UploadFile = File(...),
model_id: str = Query("fire_detection"),
confidence: float = Query(0.5),
iou: float = Query(0.45)
iou: float = Query(0.45),
algorithm_config: Optional[str] = Query(None, description="算法配置JSON字符串")
):
"""
图片检测接口
Args:
algorithm_config: 算法配置JSON，例如：
{
"enable_stationary_detection": true,
"enable_loitering_detection": false,
"stationary_threshold": 10.0,
"position_tolerance": 50,
"loitering_threshold": 300.0,
"movement_threshold": 5.0
}
"""
from main import model_service
from services.detection_service import DetectionService
detection_service = DetectionService(model_service)
# 解析算法配置
algo_config = None
if algorithm_config:
try:
algo_config = json.loads(algorithm_config)
except json.JSONDecodeError as e:
logger.warning(f"算法配置解析失败: {e}")
try:
contents = await file.read()
nparr = np.frombuffer(contents, np.uint8)
@@ -32,10 +58,14 @@ async def detect_image(
data={}
)
result = await detection_service.detect_image(frame, model_id, confidence, iou)
result = await detection_service.detect_image(
frame, model_id, confidence, iou, algorithm_config=algo_config
)
if result['success']:
annotated_frame = detection_service.draw_detections(frame, result['detections'])
annotated_frame = detection_service.draw_detections(
frame, result['detections'], algorithm_config=algo_config
)
# 将标注后的图片转换为 base64
_, buffer = cv2.imencode('.jpg', annotated_frame)
@@ -47,7 +77,9 @@ async def detect_image(
data={
"detections": result['detections'],
"image_base64": img_base64,
"stats": result['stats']
"stats": result['stats'],
"alerts": result.get('alerts', []),
"behavior_stats": result.get('behavior_stats', {})
}
)
else:
@@ -64,3 +96,66 @@ async def detect_image(
message=f"检测失败: {str(e)}",
data={}
)
@router.get("/algorithms/config")
async def get_algorithm_config():
"""获取算法配置选项"""
return {
"algorithms": [
{
"id": "stationary_detection",
"name": "静止检测",
"description": "检测人员在同一位置静止停留",
"params": [
{
"name": "stationary_threshold",
"label": "静止阈值",
"type": "number",
"default": 10.0,
"min": 1.0,
"max": 300.0,
"unit": "秒",
"description": "超过此时间视为静止"
},
{
"name": "position_tolerance",
"label": "位置容差",
"type": "number",
"default": 50,
"min": 10,
"max": 200,
"unit": "像素",
"description": "位置匹配容差范围"
}
]
},
{
"id": "loitering_detection",
"name": "徘徊检测",
"description": "检测人员长时间停留（需要跟踪ID）",
"params": [
{
"name": "loitering_threshold",
"label": "徘徊阈值",
"type": "number",
"default": 300.0,
"min": 60.0,
"max": 1800.0,
"unit": "秒",
"description": "超过此时间视为徘徊"
},
{
"name": "movement_threshold",
"label": "移动阈值",
"type": "number",
"default": 5.0,
"min": 1.0,
"max": 50.0,
"unit": "像素",
"description": "小于此移动视为静止"
}
]
}
]
}

14
apps/server/services/camera_service.py
View File

@@ -249,11 +249,21 @@ class CameraService:
logger.info(f"发送检测结果: {len(result['detections'])} 个目标, {result['stats']}")
await websocket.send_json({
detection_message = {
'type': 'detection',
'detections': result['detections'],
'stats': result['stats']
})
}
# 包含行为告警信息
if 'alerts' in result and result['alerts']:
detection_message['alerts'] = result['alerts']
logger.info(f"发送告警: {len(result['alerts'])} 个")
if 'behavior_stats' in result:
detection_message['behavior_stats'] = result['behavior_stats']
await websocket.send_json(detection_message)
_, buffer = cv2.imencode('.jpg', frame, [cv2.IMWRITE_JPEG_QUALITY, 80])
import base64

251
apps/server/services/detection_service.py
View File

@@ -7,6 +7,8 @@ import logging
from typing import Dict, List, Optional
from PIL import Image, ImageDraw, ImageFont
from .loitering_service import get_loitering_service
logger = logging.getLogger(__name__)
class DetectionService:
@@ -18,64 +20,20 @@ class DetectionService:
os.makedirs(self.results_dir, exist_ok=True)
os.makedirs(self.temp_dir, exist_ok=True)
def draw_detections(self, frame: np.ndarray, detections: List[Dict], fps: float = 0) -> np.ndarray:
try:
img_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
pil_img = Image.fromarray(img_rgb)
draw = ImageDraw.Draw(pil_img)
try:
font = ImageFont.truetype("/System/Library/Fonts/PingFang.ttc", 20)
font_large = ImageFont.truetype("/System/Library/Fonts/PingFang.ttc", 24)
except:
font = ImageFont.load_default()
font_large = font
class_colors = {
'Fire': (255, 0, 0),
'Smoke': (128, 128, 128),
'person': (0, 255, 0),
'helmet': (255, 255, 0),
'no_helmet': (255, 0, 255),
'cigarette': (0, 165, 255) # 橙色，用于抽烟检测
}
for det in detections:
x1, y1, x2, y2 = det['bbox']
class_name = det['class']
conf = det['confidence']
label = det['label']
color = class_colors.get(class_name, (0, 255, 0))
draw.rectangle([x1, y1, x2, y2], outline=color, width=3)
label_text = f"{label} {conf:.2f}"
bbox = draw.textbbox((0, 0), label_text, font=font)
text_w = bbox[2] - bbox[0]
text_h = bbox[3] - bbox[1]
draw.rectangle([x1, y1 - text_h - 4, x1 + text_w + 4, y1], fill=color)
draw.text((x1 + 2, y1 - text_h - 2), label_text, fill=(255, 255, 255), font=font)
if fps > 0:
fps_text = f"FPS: {fps:.1f} | Detections: {len(detections)}"
draw.text((10, 10), fps_text, fill=(0, 255, 0), font=font)
return cv2.cvtColor(np.array(pil_img), cv2.COLOR_RGB2BGR)
except Exception as e:
logger.error(f"绘制检测结果失败: {e}")
return frame
# 初始化徘徊检测服务（懒加载，实际初始化在第一次使用时）
self.loitering_service = get_loitering_service()
async def detect_image(
self,
self,
image: np.ndarray,
model_id: str,
confidence: float = 0.5,
iou: float = 0.45
iou: float = 0.45,
algorithm_config: Optional[Dict] = None
) -> Dict:
start_time = time.time()
model = await self.model_service.load_model(model_id)
if not model:
return {
@@ -84,10 +42,10 @@ class DetectionService:
'detections': [],
'stats': None
}
try:
results = model(image, conf=confidence, iou=iou, verbose=False)
detections = []
for result in results:
boxes = result.boxes
@@ -96,21 +54,21 @@ class DetectionService:
conf = float(box.conf[0].cpu().numpy())
cls = int(box.cls[0].cpu().numpy())
class_name = result.names[cls]
label_map = self.model_service.model_configs[model_id]['labels']
label = label_map.get(class_name, class_name)
detections.append({
'class': class_name,
'label': label,
'confidence': round(conf, 3),
'bbox': [int(x1), int(y1), int(x2), int(y2)]
})
processing_time = time.time() - start_time
avg_confidence = sum(d['confidence'] for d in detections) / len(detections) if detections else 0
return {
result_data = {
'success': True,
'message': '检测完成',
'detections': detections,
@@ -121,6 +79,14 @@ class DetectionService:
'model_used': model_id
}
}
# 如果启用了行为检测算法
if algorithm_config and detections:
result_data = self._apply_behavior_analysis(
result_data, algorithm_config
)
return result_data
except Exception as e:
logger.error(f"图片检测失败: {e}")
return {
@@ -186,9 +152,40 @@ class DetectionService:
}
}
# 如果是人员检测模型，进行行为分析
logger.info(f"[DetectionService] 模型: {model_id}, 检测目标: {len(detections)}")
if model_id == 'loitering_detection' and detections:
logger.info("[DetectionService] 调用行为分析...")
# 确保服务已初始化
if not self.loitering_service.is_initialized:
logger.info("[DetectionService] 初始化徘徊检测服务...")
self.loitering_service.initialize(
# 检测阈值（用于判断是否静止/徘徊）
stationary_threshold=10.0,
position_tolerance=50,
loitering_threshold=300.0,
movement_threshold=5.0,
# 告警阈值（用于触发告警，应该比检测阈值高）
stationary_alert_threshold=30.0,
loitering_alert_threshold=600.0,
# 启用告警
enable_stationary_alert=True,
enable_loitering_alert=True
)
behavior_result = self.loitering_service.process_detections(
detections,
use_tracking=False # 可以改为 True 如果使用跟踪
)
detections = behavior_result['detections']
result_data['alerts'] = behavior_result['alerts']
result_data['behavior_stats'] = behavior_result['stats']
logger.info(f"[DetectionService] 行为分析完成: alerts={len(behavior_result['alerts'])}, stats={behavior_result['stats']}")
if draw:
frame = self.draw_detections(frame, detections, fps)
return frame, result_data
except Exception as e:
logger.error(f"帧检测失败: {e}")
@@ -197,3 +194,139 @@ class DetectionService:
'detections': [],
'stats': None
}
def _apply_behavior_analysis(
self,
result_data: Dict,
algorithm_config: Dict
) -> Dict:
"""
应用行为分析算法
Args:
result_data: 检测结果
algorithm_config: 算法配置
{
"enable_stationary_detection": true,
"enable_loitering_detection": false,
"stationary_threshold": 10.0,
"position_tolerance": 50,
...
}
Returns:
添加行为分析结果的检测结果
"""
detections = result_data['detections']
# 检查是否需要行为分析
enable_stationary = algorithm_config.get('enable_stationary_detection', False)
enable_loitering = algorithm_config.get('enable_loitering_detection', False)
if not enable_stationary and not enable_loitering:
return result_data
try:
# 使用前端传入的配置初始化服务
self.loitering_service.initialize(
stationary_threshold=algorithm_config.get('stationary_threshold', 10.0),
position_tolerance=algorithm_config.get('position_tolerance', 50),
loitering_threshold=algorithm_config.get('loitering_threshold', 300.0),
movement_threshold=algorithm_config.get('movement_threshold', 5.0),
enable_stationary_alert=enable_stationary,
enable_loitering_alert=enable_loitering
)
# 处理检测
behavior_result = self.loitering_service.process_detections(
detections,
use_tracking=enable_loitering # 只有启用徘徊检测时才使用跟踪
)
result_data['detections'] = behavior_result['detections']
result_data['alerts'] = behavior_result['alerts']
result_data['behavior_stats'] = behavior_result['stats']
except Exception as e:
logger.error(f"行为分析失败: {e}")
result_data['behavior_error'] = str(e)
return result_data
def draw_detections(
self,
frame: np.ndarray,
detections: List[Dict],
fps: float = 0,
algorithm_config: Optional[Dict] = None
) -> np.ndarray:
"""
绘制检测结果和行为告警
Args:
frame: 图像帧
detections: 检测结果列表（可能包含 stationary_info/loitering_info）
fps: 帧率
algorithm_config: 算法配置（已废弃，保留用于向后兼容）
"""
try:
img_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
pil_img = Image.fromarray(img_rgb)
draw = ImageDraw.Draw(pil_img)
try:
font = ImageFont.truetype("/System/Library/Fonts/PingFang.ttc", 20)
font_large = ImageFont.truetype("/System/Library/Fonts/PingFang.ttc", 24)
except:
font = ImageFont.load_default()
font_large = font
class_colors = {
'Fire': (255, 0, 0),
'Smoke': (128, 128, 128),
'person': (0, 255, 0),
'helmet': (255, 255, 0),
'no_helmet': (255, 0, 255),
'cigarette': (0, 165, 255)
}
for det in detections:
x1, y1, x2, y2 = det['bbox']
class_name = det['class']
conf = det['confidence']
label = det['label']
# 根据是否有行为告警选择颜色
color = class_colors.get(class_name, (0, 255, 0))
# 检查行为告警
if algorithm_config:
if 'stationary_info' in det:
info = det['stationary_info']
if info.get('is_stationary'):
color = (0, 0, 255) # 红色警告
label = f"静止{int(info['duration'])}s"
if 'loitering_info' in det:
info = det['loitering_info']
if info.get('is_loitering'):
color = (255, 0, 0) # 蓝色警告
label = f"徘徊{int(info['loitering_duration']//60)}min"
draw.rectangle([x1, y1, x2, y2], outline=color, width=3)
label_text = f"{label} {conf:.2f}"
bbox = draw.textbbox((0, 0), label_text, font=font)
text_w = bbox[2] - bbox[0]
text_h = bbox[3] - bbox[1]
draw.rectangle([x1, y1 - text_h - 4, x1 + text_w + 4, y1], fill=color)
draw.text((x1 + 2, y1 - text_h - 2), label_text, fill=(255, 255, 255), font=font)
if fps > 0:
fps_text = f"FPS: {fps:.1f} | Detections: {len(detections)}"
draw.text((10, 10), fps_text, fill=(0, 255, 0), font=font)
return cv2.cvtColor(np.array(pil_img), cv2.COLOR_RGB2BGR)
except Exception as e:
logger.error(f"绘制检测结果失败: {e}")
return frame

168
apps/server/services/loitering_service.py Normal file
View File

@@ -0,0 +1,168 @@
"""
徘徊检测服务
集成行为检测算法到后端服务
"""
import sys
import os
from typing import Dict, List, Optional
import logging
# 添加算法模块路径
sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', '..', '..', 'models', 'loitering_detection'))
from processors import BehaviorProcessor
logger = logging.getLogger(__name__)
class LoiteringService:
"""
徘徊检测服务
为视频流检测提供行为分析功能：
- 静止检测（基于位置，无需跟踪）
- 徘徊检测（基于跟踪ID）
"""
def __init__(self):
self.processor = None
self.is_initialized = False
def initialize(
self,
stationary_threshold: float = 10.0,
position_tolerance: int = 50,
loitering_threshold: float = 300.0,
movement_threshold: float = 5.0,
enable_stationary_alert: bool = True,
enable_loitering_alert: bool = True,
stationary_alert_threshold: Optional[float] = None,
loitering_alert_threshold: Optional[float] = None
):
"""
初始化服务
Args:
stationary_threshold: 静止检测阈值（秒）- 用于判断是否静止
position_tolerance: 位置容差（像素）
loitering_threshold: 徘徊检测阈值（秒）- 用于判断是否徘徊
movement_threshold: 移动阈值（像素）
enable_stationary_alert: 是否启用静止告警
enable_loitering_alert: 是否启用徘徊告警
stationary_alert_threshold: 静止告警阈值（秒）- 超过此时间产生告警，默认等于 stationary_threshold
loitering_alert_threshold: 徘徊告警阈值（秒）- 超过此时间产生告警，默认等于 loitering_threshold
"""
try:
self.processor = BehaviorProcessor(
stationary_threshold=stationary_threshold,
position_tolerance=position_tolerance,
loitering_threshold=loitering_threshold,
movement_threshold=movement_threshold,
enable_stationary_alert=enable_stationary_alert,
enable_loitering_alert=enable_loitering_alert,
stationary_alert_threshold=stationary_alert_threshold if stationary_alert_threshold is not None else stationary_threshold,
loitering_alert_threshold=loitering_alert_threshold if loitering_alert_threshold is not None else loitering_threshold
)
self.is_initialized = True
logger.info(f"徘徊检测服务初始化成功: 静止阈值={stationary_threshold}s, 告警阈值={stationary_alert_threshold or stationary_threshold}s")
except Exception as e:
logger.error(f"徘徊检测服务初始化失败: {e}")
self.is_initialized = False
def process_detections(
self,
detections: List[Dict],
use_tracking: bool = False,
track_id_key: str = 'track_id'
) -> Dict:
"""
处理检测结果
Args:
detections: YOLO检测结果列表
use_tracking: 是否使用跟踪ID
track_id_key: 跟踪ID字段名
Returns:
{
'detections': 添加行为信息的检测结果,
'alerts': 触发的告警列表,
'stats': 统计信息
}
"""
if not self.is_initialized or not self.processor:
return {
'detections': detections,
'alerts': [],
'stats': {'error': '服务未初始化'}
}
try:
return self.processor.process(
detections=detections,
use_tracking=use_tracking,
track_id_key=track_id_key
)
except Exception as e:
logger.error(f"处理检测结果失败: {e}")
return {
'detections': detections,
'alerts': [],
'stats': {'error': str(e)}
}
def get_stationary_persons(self) -> List[Dict]:
"""获取所有静止人员"""
if not self.is_initialized or not self.processor:
return []
return self.processor.get_stationary_persons()
def get_loitering_persons(self) -> List[Dict]:
"""获取所有徘徊人员"""
if not self.is_initialized or not self.processor:
return []
return self.processor.get_loitering_persons()
def reset(self):
"""重置检测器"""
if self.processor:
self.processor.reset()
logger.info("徘徊检测器已重置")
def get_config(self) -> Dict:
"""获取当前配置"""
if not self.is_initialized or not self.processor:
return {'error': '服务未初始化'}
return self.processor.get_config()
def get_stats(self) -> Dict:
"""获取统计信息"""
if not self.is_initialized or not self.processor:
return {'error': '服务未初始化'}
stats = {
'stationary_count': len(self.get_stationary_persons()),
'loitering_count': len(self.get_loitering_persons()),
'config': self.get_config()
}
return stats
# 全局服务实例
_loitering_service: Optional[LoiteringService] = None
def get_loitering_service() -> LoiteringService:
"""获取全局徘徊检测服务实例"""
global _loitering_service
if _loitering_service is None:
_loitering_service = LoiteringService()
return _loitering_service
def initialize_loitering_service(**kwargs):
"""初始化全局徘徊检测服务"""
service = get_loitering_service()
service.initialize(**kwargs)
return service

1
apps/server/services/model_service.py
View File

@@ -82,7 +82,6 @@ class ModelService:
return None
if model_id in self.models:
logger.info(f"模型已加载: {model_id}")
return self.models[model_id]
config = self.model_configs[model_id]