中国农业科学院农业信息研究所 中国农学会
Agricultural Information Institute of CAAS China Association of Agricultural Science Societies
En
Ch
中国农业科学院农业信息研究所 中国农学会 主办

AI检索

订阅

返回顶部
当前位置: 首页 > 文章 > 果园自然环境下采摘机器人路径识别方法 江苏农业学报 2019,35 (5) 1222-1231
Position: Home > Articles > Path recognition method of picking robot based on orchard natural envi-ronment Jiangsu Journal of Agricultural Sciences 2019,35 (5) 1222-1231

果园自然环境下采摘机器人路径识别方法

作  者：
刘波;杨长辉;熊龙烨;王恺;王毅
单  位：
重庆理工大学机械工程学院
关键词：
果园环境;采摘机器人;道路中心线;光照条件
摘  要：
果园自然环境下光照条件是不同的,而不同的光照条件又会对采摘机器人的路径导航产生不同影响,针对此环境,本研究提出了一种根据光照度进行分类并使用不同算法提取道路中心线的方法.首先对图像亮度分量和光照度进行研究,并将光照度划分为低光照、正常光照和高光照3个等级.在低光照条件下通过分离出S通道,然后利用K-means与Ncut算法对其进行分割,在正常光照条件下采用Otsu算法对S通道进行分割,在高光照条件下则通过K-means与Ncut算法对Cg、Cb与Cr通道进行差分运算后的图像进行分割.将分割后的图像进行边缘检测并提取道路轮廓,并通过最小二乘法实现道路中心线的获取.最后选用150张不同光照条件下的图像进行了静态试验验证,并通过课题组自行研制的采摘机器人进行了动态试验验证,静态试验验证结果表明,3种光照条件下图像分割区域与道路真实区域的平均重合度为96.74%、平均分割误差为2.01%、平均道路中心线平均偏差为2.71像素,平均耗时为0.182 s;动态试验验证结果表明,3种光照条件下的平均横向偏移距离为3.1 cm.表明,该方法在不同光照条件下具有较高的精度和实时性,能够满足采摘机器人在自然光照条件下的路径识别及导航需求.
译  名：
Path recognition method of picking robot based on orchard natural envi-ronment
作  者：
LIU Bo;YANG Chang-hui;XIONG Long-ye;WANG Kai;WANG Yi;College of Mechanical Engineering, Chongqing University of Technology;School of Mechanical Engineering, Xi'an Jiaotong University;College of Mechanical Engineering, Chongqing University;
关键词：
orchard environment;;picking robot;;road centerline;;light condition
摘  要：
The light conditions are different in the natural environment of orchard, and different light conditions will have different effects on the path navigation of the picking robot. In response to this environment, a method of classifying road centerlines based on illumination intensity and extracting road centerlines using different algorithms was proposed. Firstly, the image brightness component and light intensity were studied, and the light intensity was divided into three levels: low light, normal light and high light. The S channel was separated under low light condition, then K-means and Ncut algorithm were used to segment it. The S channel was segmented by Otsu algorithm under normal illumination condition. Under the high-light condition, K-means and Ncut algorithm were used to segment images after differential operation in Cg, Cb and Cr channels. The segmented image was used for edge detection and road contour extraction, and the road center line was acquired by the least squares method. Finally, 150 images under different light conditions were selected for static experiment verification, and the dynamic experiments were verified by the picking robot developed by the research team. The static experimental verification results showed that the average coincidence degree between the image segmentation region and the real road region under the three light conditions was 96.74%, the average segmentation error was 2.01%, and the average deviation of average road centerline was 2.71 pixels, the average time was 0.182 s. The dynamic experimental verification results showed that the average lateral offset distance under three light conditions was 3.1 cm. These results indicate that the method has high precision and real-time, performance under different light condictions and can meet the path recognition and navigation requirements of the picking robot under natural light condition.

相似文章

导出引用
计量
文章访问数: 12
HTML全文浏览量: 0
PDF下载量: 0

所属期刊

推荐期刊

友情链接Friendly link
农业信息研究所Agricultural information institute 农业智能知识服务平台Agricultural intelligent knowledge service platform 农业学术服务平台Agricultural Knowledge Service System 中国知网CNKI 维普网VIP 万方WANFANG DATA
联系我们Contact us

电话Tell：010-82109652-808

邮箱E-mail：agrijournal@caas.cn

服务Service
关于我们About us 意见建议Opinions and suggest
版权说明Copyright
单位：中国农业科学院农业信息研究所 中国农学会 © 2017中国农业科学院农业信息研究所 ICP备案号：京ICP备09089781号-46