基于小波变换多阈值彩色铁谱图像的分割方法

Segmentation Method of Color Image of Ferrography Based on Multi-threshold Wavelet Transform

提出了一种基于小波变换多阈值彩色铁谱图像的分割方法。该方法首先将彩色铁谱图像分割为3个层次图像,然后对各单色图的直方图通过小波变换多阈值进行分割,形成背景色、一般磨粒和异常磨粒3个分割区域,最后再合成彩色图像,实现对背景色、正常磨粒沉积链、异常大磨粒的划分。实验证明该方法可以有效地提高所需图像的质量。

A segmentation method of color image of ferrography based on multi-threshold wavelet transform was presented.By this method,the color image of ferrography was segmented into three-level images,the histogram graphes of the monochrome images were segmented to form three separate regions,the background color,the general abrasive and abnormal abrasive based on multi-threshold wavelet transform,and another color image was synthesized which can distinguish the background color,the normal abrasive deposition chain and abnomal larger abrasive. The results of experiment show that the method can improve the quality of the image effectively.