基于图像角点匹配的机械加工零件表面缺陷检测

Surface Defect Detection of Machined Parts Based on Image Corner Matching

当前在对机械加工零件表面缺陷检测时,存在表面缺陷检测精度差的问题,导致检测结果不理想。提出一种基于图像角点匹配的机械加工零件表面缺陷检测方法,利用曲率空间检测零件图像的角点,采用泰勒级数删除伪角点。对特征点邻域梯度方向进行角度限度和就近投影,同时借助双向匹配方法进行机械加工零件图像角点匹配。在上述操作的基础上,利用一维直方图的阈值分割对零件图像进行分割处理,最终实现机械加工零件表面缺陷检测。实验结果表明,该方法能够获取高精度的零件表面缺陷检测结果,且对加工零件缺陷厚度、孔洞缺陷及缺陷最大边界距离的测量均较为准确。

At present,when detecting surface defects of machined parts,there is a problem of poor surface defect detection accuracy,which leads to unsatisfactory detection results.A surface defect detection method for machined parts based on image corner matching is proposed.The corner points of the part image are detected by the curvature space,and the pseudo corner points are deleted by using Taylor series.The angle limit and the nearest projection are performed on the gradient direction of the feature point neighborhood,and the corner points of the machined parts image are matched by the two-way matching method.On the basis of the above operation,the part image is segmented by the threshold segmentation of the one-dimensional histogram,and finally the surface defect detection of the machined part is realized.The experimental results show that the proposed method can obtain high-precision parts surface defect detection results,and the measurement of defect thickness,hole defects and maximum boundary distance of the machined parts are relatively accurate.