基于YOLOv26的智能农业监测系统

基于YOLOv26的智能农业监测系统

1. 系统架构

1.1 整体架构

本系统采用分层架构设计，主要包含以下核心模块：

• 数据采集模块：通过固定摄像头、无人机和传感器获取农业数据
• 数据预处理模块：对原始数据进行标准化、增强和地理坐标处理
• 目标检测模块：基于YOLOv26检测作物、病虫害、杂草等
• 生长分析模块：分析作物的生长状态和健康状况
• 产量预测模块：预测作物产量
• 地理信息处理模块：处理地理坐标和空间分析
• 决策支持模块：根据检测结果生成农业管理建议
• 结果可视化模块：将检测和分析结果可视化
• 部署与集成模块：实现系统在边缘设备和云端的部署，与现有农业管理系统集成

1.2 技术栈

2. 核心功能模块

2.1 数据采集模块

功能描述：

• 通过固定摄像头、无人机和传感器获取农业数据
• 支持多源数据融合，包括可见光、多光谱和热成像数据
• 实现数据的实时传输和存储

代码实现：

importcv2importnumpyasnpimporttimeimportthreadingfromdatetimeimportdatetimefrommavsdkimportSystemimportasyncioimportserialimportosclassDataCollector:def__init__(self,config):self.config=config self.cameras={}self.sensors={}self.data_queue=[]self.running=Falsedefinitialize_cameras(self):"""初始化摄像头"""forcam_id,cam_configinself.config.get('cameras',{}).items():try:cap=cv2.VideoCapture(cam_config['index'])cap.set(cv2.CAP_PROP_FRAME_WIDTH,cam_config.get('width',1280))cap.set(cv2.CAP_PROP_FRAME_HEIGHT,cam_config.get('height',720))self.cameras[cam_id]=capprint(f"Camera{cam_id}initialized")exceptExceptionase:print(f"Failed to initialize camera{cam_id}:{e}")definitialize_sensors(self):"""初始化传感器"""forsensor_id,sensor_configinself.config.get('sensors',{}).items():try:ser=serial.Serial(port=sensor_config['port'],baudrate=sensor_config.get('baudrate',9600),timeout=sensor_config.get('timeout',1))self.sensors[sensor_id]=serprint(f"Sensor{sensor_id}initialized")exceptExceptionase:print(f"Failed to initialize sensor{sensor_id}:{e}")defcapture_from_camera(self,cam_id):"""从指定摄像头捕获图像"""ifcam_idinself.cameras:cap=self.cameras[cam_id]ret,frame=cap.read()ifret:timestamp=datetime.now().isoformat()return{'frame':frame,'timestamp':timestamp,'source':f'camera_{cam_id}'}returnNonedefcapture_from_all_cameras(self):"""从所有摄像头捕获图像"""frames=[]forcam_idinself.cameras:frame_data=self.capture_from_camera(cam_id)ifframe_data:frames.append(frame_data)returnframesdefread_from_sensor(self,sensor_id):"""从指定传感器读取数据"""ifsensor_idinself.sensors:ser=self.sensors[sensor_id]try:data=ser.readline().decode('utf-8').strip()timestamp=datetime.now().isoformat()return{'data':data,'timestamp':timestamp,'source':f'sensor_{sensor_id}'}exceptExceptionase:print(f"Failed to read from sensor{sensor_id}:{e}")returnNonedefread_from_all_sensors(self):"""从所有传感器读取数据"""sensor_data=[]forsensor_idinself.sensors:data=self.read_from_sensor(sensor_id)ifdata:sensor_data.append(data)returnsensor_dataclassUAVDataCollector:def__init__(self,drone_ip="127.0.0.1",drone_port=50040):self.drone=System()self.drone_ip=drone_ip self.drone_port=drone_portasyncdefconnect(self):"""连接到无人机"""awaitself.drone.connect(system_address=f"udp://{self.drone_ip}:{self.drone_port}")print("Connected to drone")asyncdefplan_mission(self,field_coords,altitude=50):"""规划巡检任务"""# 简化实现，实际应考虑地形、风向等因素waypoints=[]forcoordinfield_coords:waypoints.append((coord[0],coord[1],altitude))returnwaypointsasyncdefexecute_mission(self,waypoints,image_save_dir):"""执行飞行任务并采集数据"""ifnotwaypoints:print("No mission planned")return[]# 起飞awaitself.drone.action.arm()awaitself.drone.action.takeoff()awaitasyncio.sleep(5)# 等待无人机稳定collected_images=[]# 按路径飞行并采集图像fori,waypointinenumerate(waypoints):print(f"Flying to waypoint{i+1}:{waypoint}")# 飞行到目标点awaitself.drone.action.goto_location(waypoint[0],waypoint[1],waypoint[2],0# 最后一个参数是偏航角)# 等待到达目标点awaitasyncio.sleep(10)# 采集图像（模拟）timestamp=datetime.now().strftime("%Y%m%d_%H%M%S")image_path=f"{image_save_dir}/uav_image_{i}_{timestamp}.jpg"# 创建模拟图像image=np.zeros((480,640,3),dtype=np.uint8)cv2.putText(image,f"Field Image{i}",(50,50),cv2.FONT_HERSHEY_SIMPLEX,1,(255,255,255),2)cv2.imwrite(image_path,image)collected_images.append(image_path)# 返航awaitself.drone.action.return_to_launch()awaitasyncio.sleep(10)awaitself.drone.action.disarm()print(f"Mission completed. Collected{len(collected_images)}images")returncollected_imagesclassDataCollectorManager:def__init__(self,config):self.config=config self.camera_collector=DataCollector(config)self.uav_collector=UAVDataCollector()self.data_storage=[]defstart_collection(self):"""启动数据采集"""self.camera_collector.initialize_cameras()self.camera_collector.initialize_sensors()self.running=True# 启动摄像头采集线程defcamera_collection_loop():whileself.running:frames=self.camera_collector.capture_from_all_cameras()forframe_datainframes:self.data_storage.append(frame_data)time.sleep(1)# 每秒采集一次# 启动传感器采集线程defsensor_collection_loop():whileself.running:sensor_data=self.camera_collector.read_from_all_sensors()fordatainsensor_data:self.data_storage.append(data)time.sleep(5)# 每5秒采集一次self.camera_thread=threading.Thread(target=camera_collection_loop)self.camera_thread.daemon=Trueself.camera_thread.start()self.sensor_thread=threading.Thread(target=sensor_collection_loop)self.sensor_thread.daemon=Trueself.sensor_thread.start()print("Data collection started")defstop_collection(self):"""停止数据采集"""self.running=Falseifhasattr(self,'camera_thread'):self.camera_thread.join(timeout=2)ifhasattr(self,'sensor_thread'):self.sensor_thread.join(timeout=2)# 释放摄像头资源forcam_id,capinself.camera_collector.cameras.items():try:cap.release()except:pass# 关闭传感器forsensor_id,serinself.camera_collector.sensors.items():try:ser.close()except:passprint("Data collection stopped")asyncdefrun_uav_inspection(self,field_coords,image_save_dir):"""运行无人机巡检"""awaitself.uav_collector.connect()waypoints=awaitself.uav_collector.plan_mission(field_coords)returnawaitself.uav_collector.execute_mission(waypoints,image_save_dir)# 示例用法config={'cameras':{'field1':{'index':0,'width':1280,'height':720},'field2':{'index':1,'width':1280,'height':720}},'sensors':{'soil_moisture':{'port':'COM3','baudrate':9600},'temperature':{'port':'COM4','baudrate':9600}}}manager=DataCollectorManager(config)manager.start_collection()# 运行无人机巡检（异步）asyncdefrun_uav_inspection():field_coords=[(39.9042,116.4074),(39.9052,116.4074),(39.9052,116.4084),(39.9042,116.4084)]images=awaitmanager.run_uav_inspection(field_coords,'data/uav_images')returnimages# 运行10秒后停止采集time.sleep(10)manager.stop_collection()print(f"Collected{len(manager.data_storage)}data points")