基于视觉协同的双机械臂物品动态抓取系统设计与实现  

Design and Realization of Dual Robotic Arm Item Dynamic Gripping System Based on Visual Collaboration

作　　者：赵源麒 王清珍[1] 

机构地区：[1]郑州科技学院大数据与人工智能学院,河南 郑州

出　　处：《仪器与设备》2024年第2期107-116,共10页Instrumentation and Equipments

摘　　要：本文旨在设计和实现基于视觉协同的双机械臂物品动态抓取系统,采用深度相机和普通相机实现对目标物的视觉感知。搭建系统,实现空间坐标转换,将像素坐标转换为空间坐标,为后续的动态抓取路径规划提供准确的位置信息。在视觉协同算法方面,设计了动态抓取路径规划算法,以实现高效而精准的物品抓取;同时,采用视觉反馈控制算法,实现对抓取过程中的实时调整和精细控制。实验结果表明,色域值实验验证了系统的视觉感知准确性,动态抓取精度测试和响应速度测试均证实了系统的高效性和稳定性。通过本研究,得出所设计的基于视觉协同的双机械臂物品动态抓取系统在实现精准抓取和高效响应速度方面取得了显著成果,为工业自动化领域的应用提供了有力支持。The purpose of this paper is to design and implement a dual robotic arm item dynamic grasping system based on visual collaboration, using depth cameras and ordinary cameras to realize visual perception of the target item. The system is built to realize spatial coordinate conversion, which converts pixel coordinates to spatial coordinates to provide accurate position information for subsequent dynamic grasping path planning. In terms of visual collaborative algorithms, a dynamic grasping path planning algorithm is designed to achieve efficient and accurate item grasping;meanwhile, a visual feedback control algorithm is used to realize real-time adjustment and fine control of the grasping process. The experimental results show that the color gamut value experiment verifies the visual perception accuracy of the system, and both the dynamic grasping accuracy test and the response speed test confirm the high efficiency and stability of the system. Through this study, it is concluded that the designed dual robotic arm item dynamic grasping system based on visual synergy has achieved remarkable results in realizing accurate grasping and efficient response speed, which provides strong support for applications in industrial automation.

关 键 词：视觉协同 双机械臂 动态抓取 深度相机 实时控制 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]