针对采摘机器人自主行走导航过程中,难以准确定位其与果树之间的相对位置,难以准确估计果树树干姿态的问题,提出基于双目eye in hand系统的多角度树干位姿估计方法。利用YOLOv5深度学习方法与半全局块匹配算法识别树干并生成局部点云;...针对采摘机器人自主行走导航过程中,难以准确定位其与果树之间的相对位置,难以准确估计果树树干姿态的问题,提出基于双目eye in hand系统的多角度树干位姿估计方法。利用YOLOv5深度学习方法与半全局块匹配算法识别树干并生成局部点云;利用半径滤波和体素滤波减少树干点云数据;利用闭环式手眼标定方法对双目eye in hand系统进行标定,并对同一树干多角度相机位置的点云数据进行拼接;利用随机抽样一致(RANSAC)算法与无约束最小二乘法估计并优化树干的位置和姿态,获取树干的圆柱体参数。通过对30幅标定板图像进行实验,闭环式手眼标定方法的平均欧式误差为3.7177 mm;采用半径滤波和体素滤波可减少98.470%的点云数据;采用RANSAC算法、圆柱体估计算法拟合树干点云数据,得到圆柱体的半径r=41.2771 mm,R_(MAE)=2.57156 mm,R_(RMSE)=2.98936 mm;无约束最小二乘法优化后r=39.4028 mm,R_(MAE)=1.98955 mm,R_(RMSE)=2.46588 mm。该文通过对双目eye in hand系统进行标定,建立坐标系转换关系,多角度采集环境信息,准确定位机器人与果树之间的相对位置,估计果树树干的姿态。展开更多
This paper discusses the coordination process for a robot gripper to approach a movingobject with feedback from an uncalibrated visual system. The dynamic of the whole system, includingtarget's random motion and t...This paper discusses the coordination process for a robot gripper to approach a movingobject with feedback from an uncalibrated visual system. The dynamic of the whole system, includingtarget's random motion and the gripper's tracking motion, is bounded to a 2-D working plane. Acamera, whose relations with the robot system and the 2-D working plane are unknown to the robotcontroller, is fixed aside to observe the object and gripper positions continually. Thus the movementsof the robot gripper can be decided on the positions of the object observed in each visual samplingmoment. The coordination of the vision and robot system is to be shown independently from therelations between the robot and the vision system, which should always be calibrated a prior forthe control of traditional robot/vision coordination system. Simulations are provided to show theproperty of the proposed method.展开更多
为提高汽车工艺涂胶质量及机械臂作业效率,针对基于深度学习的双目视觉车顶焊缝涂胶机械臂系统,提出了一种SEmYOLOv5算法,在主干网络上增加SE(squeeze and excitation)注意力机制,同时在颈部网络上增加一组采样模块,提高焊缝的识别能力...为提高汽车工艺涂胶质量及机械臂作业效率,针对基于深度学习的双目视觉车顶焊缝涂胶机械臂系统,提出了一种SEmYOLOv5算法,在主干网络上增加SE(squeeze and excitation)注意力机制,同时在颈部网络上增加一组采样模块,提高焊缝的识别能力。对提取到的图像进行图像处理,使得更好的提取车顶焊缝的特征信息从而得到特征点坐标,采用B样条曲线法对机械臂进行轨迹规划。改进后的算法相较原YOLOv5算法的mAP值提升了6.76%,针对该系统进行实验并验证了提出的基于深度学习的双目视觉车顶焊缝涂胶机械臂系统的有效性。展开更多
为解决人工切割大型铸件冒口对人体损害大,生产效率低和切割平面粗糙的问题,对视觉检测技术和机器人切割轨迹规划进行研究。首先,利用3D工业相机采集铸件三维场景点云信息;然后,通过提出的三点模板点云配准(three-point template point ...为解决人工切割大型铸件冒口对人体损害大,生产效率低和切割平面粗糙的问题,对视觉检测技术和机器人切割轨迹规划进行研究。首先,利用3D工业相机采集铸件三维场景点云信息;然后,通过提出的三点模板点云配准(three-point template point cloud registration, TTPCR)方法获取铸件切割点位的位姿信息,利用手眼标定变换矩阵把切割点位的信息变换到机械臂的基坐标系下;最后,利用空间圆弧的姿态插补求出切割轨迹的位姿信息,并用RoboDK软件开展实验。实验结果表明切割的误差小于1.3 mm,相对于传统的人工切割方法,切割豁口缝隙减少了37.5%,切割表面粗糙度降低了70.8%,切割表面平均粗糙深度降低了65.6%,满足铸件切割工艺要求,具有一定的工业应用价值。展开更多
文摘针对采摘机器人自主行走导航过程中,难以准确定位其与果树之间的相对位置,难以准确估计果树树干姿态的问题,提出基于双目eye in hand系统的多角度树干位姿估计方法。利用YOLOv5深度学习方法与半全局块匹配算法识别树干并生成局部点云;利用半径滤波和体素滤波减少树干点云数据;利用闭环式手眼标定方法对双目eye in hand系统进行标定,并对同一树干多角度相机位置的点云数据进行拼接;利用随机抽样一致(RANSAC)算法与无约束最小二乘法估计并优化树干的位置和姿态,获取树干的圆柱体参数。通过对30幅标定板图像进行实验,闭环式手眼标定方法的平均欧式误差为3.7177 mm;采用半径滤波和体素滤波可减少98.470%的点云数据;采用RANSAC算法、圆柱体估计算法拟合树干点云数据,得到圆柱体的半径r=41.2771 mm,R_(MAE)=2.57156 mm,R_(RMSE)=2.98936 mm;无约束最小二乘法优化后r=39.4028 mm,R_(MAE)=1.98955 mm,R_(RMSE)=2.46588 mm。该文通过对双目eye in hand系统进行标定,建立坐标系转换关系,多角度采集环境信息,准确定位机器人与果树之间的相对位置,估计果树树干的姿态。
文摘This paper discusses the coordination process for a robot gripper to approach a movingobject with feedback from an uncalibrated visual system. The dynamic of the whole system, includingtarget's random motion and the gripper's tracking motion, is bounded to a 2-D working plane. Acamera, whose relations with the robot system and the 2-D working plane are unknown to the robotcontroller, is fixed aside to observe the object and gripper positions continually. Thus the movementsof the robot gripper can be decided on the positions of the object observed in each visual samplingmoment. The coordination of the vision and robot system is to be shown independently from therelations between the robot and the vision system, which should always be calibrated a prior forthe control of traditional robot/vision coordination system. Simulations are provided to show theproperty of the proposed method.
文摘为提高汽车工艺涂胶质量及机械臂作业效率,针对基于深度学习的双目视觉车顶焊缝涂胶机械臂系统,提出了一种SEmYOLOv5算法,在主干网络上增加SE(squeeze and excitation)注意力机制,同时在颈部网络上增加一组采样模块,提高焊缝的识别能力。对提取到的图像进行图像处理,使得更好的提取车顶焊缝的特征信息从而得到特征点坐标,采用B样条曲线法对机械臂进行轨迹规划。改进后的算法相较原YOLOv5算法的mAP值提升了6.76%,针对该系统进行实验并验证了提出的基于深度学习的双目视觉车顶焊缝涂胶机械臂系统的有效性。