一种基于小波变换去噪的活塞环图像检测技术

A Technology of Image Measurement and Surface Defect Recognition of Piston Ring Based on Wavelet Transform

为精确测量活塞环的尺寸和识别破损缺陷,提出一种关于活塞环图像测量及破损缺陷的检测算法.该算法首先对活塞环原图像进行预处理,获取缺陷识别模型所需图像,然后通过小波变换获取清晰的活塞环图像边缘,再通过最小二乘法对活塞环图像边缘进行拟合,得到活塞环边缘曲线,建立算法模型,识别活塞环破损缺陷,得到活塞环的尺寸参数.该检测算法的缺陷识别率达99.5%,且实现活塞环的宽度亚像素测量.

In the paper,wavelet transform method is used to calculate accurate edge of piston ring image,then the least squares method is applied to fit a relative datum curve of the edge of the piston ring.Thus dimension parameters of piston ring can be calculated after calibration of pixel.Based on a defect recognition algorithm model and few image procedures,the algorithm can measure the width of piston rings,recognize piston ring breakages part and judge whether a piston ring is conforming or not.This application of detection algorithm can be used to actualize automatic online inspection by CCDs.