作　　者： ; ; ;

机构地区： 广东工业大学机电工程学院

出　　处： 《机床与液压》 2006年第4期154-155,158,共3页

摘　　要： 工件在传输工程中不可避免地发生旋转和平移,快速准确地将检测工件与标准工件图像中的目标对准是应用的关键。为此,本文研究了应用机械视觉进行工件表面质量检测的方法,提出了以工件目标的角点为识别特征,结合遗传算法快速确定变换参数的目标对准算法,在此基础上建立了一个应用机器视觉技术的工件表面质量实时检测系统。检测结果表明,该方法能有效地从各种旋转、平移后的待检测图像中实时检测出缺陷工件。 It is inevitable for workpieces to rotate or （and） translate somewhat in the process of transmission. The key for applying machine mision to test the quality of workpiece surface is to collimate object in image of testing piece with the standard piece. The quality testing of workpiece surface based on machine vision was studied. The comer-points of object were selected as set of characterpoints to match collimate object and genetic algorithm was applied to find the collimating parameter. A system of quality testing of workpiece surface based on machine vision was developed. Result shows that this system can find the defect workplece real time from the images for testing after rotating and translating.

关 键 词： 机器视觉 遗传算法 工件表面 质量检测

领　　域： [自动化与计算机技术] [自动化与计算机技术]