摘要
为实现温室中基于单目视觉的机器人自适应导航,运用Susan算子点检测的方法,融合Hough变换,提出了温室环境中的路径导航线的生成算法,并将标准Susan算子改造成适应与Hough变换的二值化形式。通过对温室中113幅图像进行实验检测可知,采用此方法可有效地检测机器人导航路径,且路径检测的识别率为94%,平均导航角为0.192 2°。
Point detecting of Susan operator were combined into Hough transformation in order to realize self-adoptive navigation for robot in greenhouse,this method was used for detecting navigation line in greenhouse.Standard Susan operator was modified into brinarization to subsequent Hough transformation.113 images in greenhouse were tested.The result of experiments show this method can detect the navigation line effectively,and the recognition rate of path detecting is 94%,average navigation angle is 0.192 2 degree.
出处
《农机化研究》
北大核心
2011年第3期151-154,共4页
Journal of Agricultural Mechanization Research
基金
国家自然科学基金项目(30571240)
国家高技术研究发展计划项目(2008AA100905
2006AA10A305-3)
江苏省自然科学基金项目(BK2008229)
