摘要
提出了一种基于机器视觉的、适用于林间行走机器人导航的路径生成算法,即采用扫描图像获得视野两边的树干底部与地面的交界点,计算得到交点中心的一系列离散点簇,通过最小二乘法拟合该中点生成导航直线的算法。Matlab仿真结果表明:该算法可有效地处理多种林间复杂环境,符合人眼视觉所识别的导航线,在图像批量处理时具有较高的可靠性和稳定性。
An algorithm of producing navigation line for robot in forestry environment based on machine vision was developed. By scanning an image, boundaries between the tree bottom and earth are recorded, and the central points between both tree bottom sides are gained and a straight line generated by least-square fitting to perform as the navigating line. A central line was obtained as the navigation guidance line. Matlab simulation results show the arithmetic efficiency in various forestry environments, in accordance with human vision recognition. The reliability and stability are demonstrated during batch processing.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2009年第7期176-179,共4页
Transactions of the Chinese Society for Agricultural Machinery
基金
国家"863"高技术研究发展计划资助项目(2006AA10Z255)
浙江省重中之重学科开放基金资助项目(56310202005)
关键词
机器人
机器视觉
路径导航
林地环境
Robot, Machine vision, Path navigation, Forestry environment
