基于Kinect V2图像传感设备的智能喷漆系统  

Intelligent painting system based on Kinect V2 image sensing equipment

作　　者：林森 李文涛[1,2] 丁乃达[1,2] 唐延东[1,2] LIN Sen;LI Wentao;DING Naida;TANG Yandong(State Key Laboratory of Robotics,Shenyang Institute of Automation,Institutes for Robotics and Intelligent Manufacturing,Chinese Academy of Sciences,Shenyang 110016,China;Institutes for Robotics and Intelligent Manufacturing,Chinese Academy of Sciences,Shenyang 110016,China;School of Electronic and Information Engineering,Liaoning Technical University,Huludao 125105,China)

机构地区：[1]中国科学院沈阳自动化研究所机器人学国家重点实验室,辽宁沈阳110016 [2]中国科学院机器人与智能制造创新研究院,辽宁沈阳110016 [3]辽宁工程技术大学电子与信息工程学院,辽宁葫芦岛125105

出　　处：《传感器与微系统》2021年第7期95-98,共4页Transducer and Microsystem Technologies

基　　金：辽宁省自然科学基金面上基金资助项目(2015020100);国家自然科学基金资助项目(91648118,61473280)。

摘　　要：针对传统喷漆机器人工作效率较低等问题,设计了一种基于Kinect V2图像传感设备的智能喷漆系统。首先,通过加入Sampson算子的RANSAC算法剔除错误匹配点对以获取精确的喷漆区域点云数据;然后,针对不规则的车漆表面,提出采用立体包围盒法寻找两点集中的最近点,缩短搜索时间,从而快速生成点云切片;最后,估算数据点法矢量,形成汽车划痕区的喷漆轨迹。实验结果表明,所设计的系统能够准确高效地完成汽车车身的漆面修复,具有良好的应用前景。Aiming at the problem of low efficiency of traditional painting robots,a smart painting system based on Kinect V2 image sensing equipment is designed.Firstly,the mismatched point pairs are eliminated by the RANSAC algorithm with adding Sampson operator to obtain accurate point cloud data of the painting area.Then,for irregular painting surfaces,the stereo bounding box method is proposed to find the nearest points of the two point sets,shortening searching time,therebythe point cloud slices are generated quickly.Finally,the normal vector of the data points are estimated to form the painting trajectory of the car scratch zone.The experimental results show that the system can complete the repair of the painting surfaces accurately and efficiently,as well as has a good application prospect.

关 键 词：Kinect V2图像传感设备 点云匹配 点云切片 喷漆轨迹 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]