基于小波及纹理特征的轴承防尘盖缺陷检测

Bearing Shield Defect Detection Based on Wavelet and Texture Feature

针对常规阈值分割算法难以检测轴承防尘盖表面较浅缺陷问题,提出一种边缘检测算法。该算法基于小波变换对防尘盖区域随机纹理的削弱作用,对分解后得到的低频部分和高频部分分别采用不同参数的Canny算子进行边缘检测。分析防尘盖区域的纹理特征与所选参数之间的关系,设计一种以防尘盖区域直方图熵值为参考的自适应策略。实验结果表明,该算法能有效分离出阈值分割难以检测的缺陷,且不易受防尘盖本身纹理的影响,为缺陷的在线检测提供了良好的支持。

Aiming at the problem that thresholding is difficult to detect the shallow defects of bearing shield surface,this paper proposes a novel edge detection algorithm.The algorithm is based on the effect of wavelet transform,which weakens the random texture on the shield region.To detect the edge,Canny operators with different parameters are separately used in the low frequency and the high frequency parts which are generated by wavelet decomposition.The relationship between the chosen parameters and the texture features of the shield region is analysed,and accordingly,an adaptive strategy is designed,which is based on the histogram＇s entropy of the shield region.Experimental results show that this algorithm can effectively extract the defects that it is difficult to detect by thresholding,and can avoid the effect of the shield＇s texture.It provides a good support to detect the defects online.