机械制造与自动化Issue(5)：188-193,6.DOI:10.19344/j.cnki.issn1671-5276.2017.05.049
基于立体视觉的工件识别与定位系统研究
Research on Parts' Recognition and Orientation Algorithm Based on Stereo Vision
摘要
Abstract
Parts' recognition and orientation system is constructed based on stereo vision, which is used to recognize and orient the parts with the weak texture target in production process. The recognition algorithm is researched on based on improved shape de-scriptor, contour moment feature and sub contour feature for increasing the accuracy of the parts' recognition. And then, the 3D po-sition of key features is computed under the constraints of epipolar geometry based on the principle of binocular distance measure-ment. OpenCV library is used with robot control system to complete the software development and by the experiment, the feasibility of recognition and orientation algorithm is proved.关键词
立体视觉/工件识别/三维姿态估计/轮廓匹配Key words
stereo vision/parts' recognition/three-dimensional pose estimating/contour feature matching分类
信息技术与安全科学引用本文复制引用
兰佳,叶文华..基于立体视觉的工件识别与定位系统研究[J].机械制造与自动化,2017,(5):188-193,6.基金项目
江苏省重大科技成果转化项目( BA2013053) （ BA2013053）
