基于立体视觉的水果采摘机器人系统调研报告摘要 : 基于立体视觉建立了水果采摘机器人系统
在图像空间利用Hough变换检测出果实目标, 并利用随机采样目标上均匀分布多个点的三维坐标信息重建果实球模型, 进而获得目标质心的空间位置坐标;通过最小二乘法研究了采摘机器人手眼标定;分析了采摘机器人的轨迹规划
实验结果表明 , 设计的自动采摘系统可以有效地消除遮挡以及立体视觉匹配失效等因素的影响 , 目标定位误差小于8mm,显著地提高了抓取的精度和可靠性
关键词 : 采摘机器人立体视觉手眼标定识别Research Report of fruit picking robot system based on stereo vision Abstract: A fruit picking robot system based on stereo vision is established
In the image space by using Hough transform to detect target fruit, and the random sampling of multiple point reconstruction of 3D coordinate information fruit ball model are uniformly distributed target, and then obtain the spatial position coordinates of the target centroid; by the least squares method of picking robot hand eye calibration; analysis of the picking robot trajectory planning
The experime