基于相位测量偏折算法的油漆缺陷检测系统的应用

Application of Coating Defect Detection System Based on Phase Measuring Deflectometry

针对机器视觉难以对汽车涂装漆膜进行缺陷检测等问题,提出了一种基于相位测量偏折算法的缺陷检测系统。该视觉系统以LCD屏幕作为结构光源,通过LCD屏幕投射到被测表面,使用4组相机捕捉结构光的反射图案,计算反射率及曲率,通过AI分析检出缺陷及缺陷种类。测试结果表明:基于相位测量偏折算法的视觉系统,可高效检测出汽车漆膜表面上的所有类型的缺陷,精度为0.2 mm,并能进行三维建模,有效满足对汽车漆膜表面上多种缺陷进行高精度、高效率检测的要求。

For the difficulty of automobile coating defect detection by machine vision,a defect detection system based on phase measurement deflectometry(PMD)algorithm was proposed.LCD screen was used as the structural light source in this vision system,and projected onto the tested surface.Four groups of cameras were used to capture the reflection pattern of structural light.The defects and defect types were analyzed and detected by AI,which were calculated by reflectivity and curvature.The test results showed that all types of defect on the surface of automobile coating with an accuracy of 0.2 mm could be detected effectively by the vision system based on PMD algorithm,and 3D modeling could be carried out also.The requirements of high-precision and high-efficiency detection of various defects on automobile coating surface could be meeted effectively.