结构不规则工件定位与抓取研究  

The Research on Orientation and Scraping of Workpiece with Asymmetry Structure

作　　者：王丹[1] 钱炜[1] 秦小健 漆思杰 顾自明 WANG Dan QIAN Wei QIN Xiaojian QI Sijie GU Ziming(School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China Yutaka Electronics （SH） CO. , LTD, Shanghai 201611, China)

机构地区：[1]上海理工大学机械工程学院,上海200093 [2]丰汉电子(上海)有限公司,上海201611

出　　处：《电子科技》2017年第1期139-142,146,共5页Electronic Science and Technology

基　　金：国家科技支撑计划基金资助项目(2015BAK16B04)

摘　　要：为解决结构不对称、质量分布不均匀的工件定位抓取问题,提出了利用工件表面特征的位置信息进行定位的方法。识别和提取工件上表面定位孔的图像坐标,联合机器人手眼标定参数获取工件的位姿信息。文中以散热器壳为例搭建了一种单目视觉系统,对机器人定位抓取系统中工件存在判定、图像分割、工件提取以及工件定位进行了研究,并通过实验进行验证,实验结果表明,该方法定位成功率高达0.92,具有较强的实用性。To solving the problem of location and grasping workpiece that with nonuniform mass distribution and asymmetry structure, a new method of orientation based on object surface features is proposed. Recognition and ex- traction the image coordinate of locating holes in the surface of workpiece, then combined image coordinate with robot EYE - TO - HAND system calibrate parameters to obtain pose information. The paper experiment on radiator case and built a whole monocular computer vision system, and did some research into robot orientation and scraping system about judgment on existence of workpiece, division of image, workpiece feature extraction and positioning. The re- suhs show that the success rate is 0.92 and the positioning method have a strong practical value.

关 键 词：单目视觉 工件定位 存在判定 图像分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]