基于图像增强的OLVF反射光磨粒图像分割算法

A Segmentation Algorithm of OLVF Reflected Wear Debris Image Based on Image Enhancement

针对反射光磨粒图像中亮度过高和过暗部分难以识别的问题,提出一种基于图像增强的OLVF(On-Line Visual Ferrograph)反射光磨粒图像分割算法。对原始图像进行背景减操作,大致区分出磨粒部分;利用形态黑帽操作增强反射光磨粒图像,使图像总体亮度均衡。对增强后的图像应用Canny边缘检测操作分割出磨粒,利用Otsu算法获取H-minima校正图像的阈值以消除局部极值干扰。最后,运用膨胀腐蚀开闭操作填充孔洞,实现了OLVF反射光磨粒图像分割,获得了准确、连续的磨粒图像。与其他同类算法相比,该算法有效抑制了反射光影响,更好地保留磨粒图像信息。

Aiming at the problem of too high brightness and too dark part in reflective wear debris image which is difficult to recognize,a segmentation algorithm of reflective wear debris image based on image enhancement of OLVF(On-Line Visual Ferrograph)was proposed.The background subtraction operation was performed on the original image to roughly distinguish the wear particles;the morphological black cap operation was used to enhance the reflected light wear particles image to make the overall brightness of the image balanced.Canny edge detection was used to segment the wear particles in the enhanced image,and Otsu algorithm was used to obtain the threshold of H-minima correction image to eliminate the interference of local extremum.The hole was filled by the opening and closing operation of expansion corrosion to realize the segmentation of wear particle image by OLVF reflection light,and the accurate and continuous wear particle image was obtained.Compared with other similar algorithms,this algorithm can effectively suppress the influence of reflected light and better retain the image information of wear particles.