基于机器视觉的滑动轴承缺陷检测系统设计

Design of Detection System for Slide Bearing Defects Based on Machine Vision

针对滑动轴承生产过程中存在的检测速度慢、自动化水平低、检测精度低等问题,提出一种基于机器视觉的滑动轴承内表面缺陷检测系统的设计方案,实现了轴承体的自动检测功能。首先设计了一种实验检测平台用于获取滑动轴承内表面的图像;通过基于形状的模板匹配算法对预处理后的图像进行匹配,实现对目标物体与缺陷区域的快速定位;为了实现对缺陷的提取,提出了一种基于区域灰度值的图像分割方法与基于区域形态学处理的特征提取方法。实验表明,系统的检测效果与传统的检测方法相比,具有明显的优越性,为滑动轴承内表面检测提供了新的方法。

For bearings production process in the slow speed,low levels of automation and accuracy problems is presented bearing inner surface defects based on machine vision detecting system design,realized the auto-detection of the bearing body. Firstly,a testing platform is designed for obtaining images of bearing inner surface; By template matching algorithm based on shape matches the image after preprocessing to achieve rapid positioning of the objects with the defective area; In order to realize the defect extraction,this paper proposes a image segmentation method based on region grey value and feature extraction method based on morphological processing area. Experiments show that performance compared with the traditional method of the system has obvious advantages,provided a new method for slide bearing inner surface defects detection.