基于计算机视觉与多传感融合的农业机械自动导航方法研究OA

To solve the problems of low navigation accuracy,unstable path tracking,and insufficient environmental per-ception ability of traditional agricultural machinery in unstructured farmland environments,this paper constructs an auto-matic navigation system for agricultural machinery based on computer vision,and combines multi-source sensing infor-mation(IMU,GPS,LiDAR data)to achieve visual recognition of work paths,inter row boundaries,and obstacles through image preprocessing,object detection,and feature extraction.Based on fuzzy PID and model predictive control(MPC)strategies,the steering angle and driving speed are dynamically adjusted to minimize path deviation and improve operational stability.