基于机器视觉的汽车精密零件缺陷检测

本文针对汽车精密零件外观质量检测的需求,利用机器视觉系统采集零件外观图像,建立数据平台。基于中值滤波和图像复杂度,提出了改进OTSU算法提取图像的外观缺陷特征,与样本库进行对比,以识别可能的外观缺陷,提高缺陷检测效率。并且,针对缺陷类型分析其影响因素,从而找到产生缺陷的原因,为零件加工过程的改善提供理论依据。同时,搭建了缺陷视觉检测平台,以汽车转向器内部套筒为对象,实现了精密零件外观缺陷在线缺陷检测。

Research on Visual Inspection of Appearance Defects of Automotive Precision Parts

In this paper, aiming at the requirements of the appearance quality inspection of automotive precision parts, the machine vision system is used to collect the appearance images of the parts and establish a data platform. Based on the median filtering and image complexity, the improved OTSU algorithm is used to extract the appearance defect features of the image, and compared with the sample library to identify possible appearance defects and improve the defect detection efficiency. Moreover, the influencing factors are analyzed for the defect type, so as to find the cause of the defect and provide a theoretical basis for the improvement of the part processing process. At the same time, the defect visual inspection platform was built, and the inner sleeve of the automobile steering gear was taken as the object to realize the online defect detection of the appearance defect of the precision parts.