Scholarlink

Abstract <p>In order to obtain more accurate robot arm target detection and capture results, a robot arm target detection and capture algorithm based on binocular vision is proposed. Give priority to binocular vision calibration, determine camera parameters, and carry out hand eye calibration for robot arm and camera. Then the binocular vision is introduced into the robot arm target detection, and the target to be detected is detected. The combination of SAC-IA coarse registration and ICP fine registration is used to complete the pose estimation of target objects based on point cloud registration. GPD algorithm is used to sample the preprocessed point cloud information to obtain candidate grasping pose. The candidate grasping pose is classified to obtain effective grasping pose. Finally, the grasping pose is determined through geometric similarity clustering to complete the robot arm target grasping. The experimental findings indicate that the suggested algorithm can effectively improve the accuracy of the robot arm target detection and capture results.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Target Detection and Grasping Algorithm of Robotic Arm Based on Binocular Vision

  • 期刊论文

  • 发表日期：

    2025年4月1日

查看全文
  • Jun Zhang

摘要

Abstract

In order to obtain more accurate robot arm target detection and capture results, a robot arm target detection and capture algorithm based on binocular vision is proposed. Give priority to binocular vision calibration, determine camera parameters, and carry out hand eye calibration for robot arm and camera. Then the binocular vision is introduced into the robot arm target detection, and the target to be detected is detected. The combination of SAC-IA coarse registration and ICP fine registration is used to complete the pose estimation of target objects based on point cloud registration. GPD algorithm is used to sample the preprocessed point cloud information to obtain candidate grasping pose. The candidate grasping pose is classified to obtain effective grasping pose. Finally, the grasping pose is determined through geometric similarity clustering to complete the robot arm target grasping. The experimental findings indicate that the suggested algorithm can effectively improve the accuracy of the robot arm target detection and capture results.