中国农机化学报2026，Vol.47Issue(2)：196-201,209,7.DOI:10.13733/j.jcam.issn.2095-5553.2026.02.026
基于单目视觉里程计的甘蔗收割导航线提取研究
Research on sugarcane harvesting navigation line extraction based on monocular visual odometry
摘要
Abstract
To develop a self-navigation system suitable for sugarcane harvesters,a method for extracting navigation lines of sugarcane harvesters based on monocular visual odometry is proposed.The moving images of the harvester were collected by using a monocular camera,and the three-dimensional point cloud information of the sugarcane field terrain was extracted by utilizing a monocular visual odometry.Subsequently,the positions of sugarcane stem lines and sugarcane ridge surfaces were extracted respectively through the RANSAC algorithm,with their intersection serving as an estimation point for the sugarcane root position.Finally,linear regression was used to fit the straight lines of the harvesting navigation lines.Experimental results demonstrate that the monocular visual odometry can track the pose of the image capture vehicle at a speed of 25 frames per second and refresh the three-dimensional point cloud at a rate of 3 frames per second.The maximum angular deviation between the navigation line fitted by this method and the manually extracted navigation line is 3.933°,and the maximum distance from the sugarcane root to the navigation line is 0.164 m.The maximum processing time per point cloud frame is 127.4 ms.关键词
甘蔗收割机/单目视觉里程计/导航线提取/甘蔗行/点云Key words
sugarcane harvester/monocular visual odometry/navigation line extraction/sugarcane field/point cloud分类
农业科技引用本文复制引用
Li Longqin,Lu Jingping,Zhang Yongxu,Feng Wukai,Lin Yundong..基于单目视觉里程计的甘蔗收割导航线提取研究[J].中国农机化学报,2026,47(2):196-201,209,7.基金项目
国家自然科学基金资助项目(51965003) （51965003）
