水稻插秧机路径追踪设计——基于SOPC嵌入式视觉导航  

Path Tracking Design of Rice Transplanter——Based on the Embedded Visual Navigation of SOPC

作　　者：谢忠华[1] 

机构地区：[1]武汉职业技术学院,武汉430074

出　　处：《农机化研究》2017年第10期213-217,共5页Journal of Agricultural Mechanization Research

基　　金：湖北省自然科学基金项目(2014CFC1079)

摘　　要：针对水稻插秧机路径追踪问题,利用SOPC嵌入式和视觉导航技术,以FPGA为硬件核心,利用3×3窗口模块、中值滤波、Sobel边缘检测和Hough变换进行移动路径图像信息处理,并结合齐次变换法将图像位姿信息转换到空间坐标系中,为插秧机进行路径导航。同时,通过人工设置标志线对水稻插秧机进行路径导航和追踪试验,结果表明:该系统在一定干扰环境下,水稻插秧机移动路径最大偏差为18cm,平均偏差为6.9m,基本达到预期目标,实现了路径追踪功能,而且可以根据偏差自动进行调整,具有精准度高、灵活性强等优点,应用前景非常广阔。In view of the rice transplanter path tracking problem, by using SOPC embedded and vision navigation tech-nology, using FPGA as the core hardware using 3 x 3 window module, median filtering, Sobel edge detection and Hough transform mobile path image information processing, and combining with the homogeneous transformation will pose infor-mation of image conversion to space coordinates to rice transplanter navigation path. Through the manual set the flag line on the rice transplanter path navigation and tracking test. Test results show that the system under interference environ-ment of rice planting machine moving path of maximum deviation of 18 cm, the average deviation is 6. 9 m, basically reached the expected target, the realization of the path tracking function, and according to deviation is automatically ad-justed with precision high, flexibility is strong wait for an advantage and application prospect is very broad.

关 键 词：水稻插秧机 路径追踪 视觉导航 SOPC嵌入式 HOUGH变换 

分 类 号：S233.91[农业科学—农业机械化工程] TP242[农业科学—农业工程]