数学形态学梯度算子检测铸铁金相图像边缘  

Mathematical morphological gradient algorithm for detecting castiron image edge

作　　者：翟改霞[1] 王春光[1] 张海军[1] 

机构地区：[1]内蒙古农业大学机电工程学院,内蒙古呼和浩特010018

出　　处：《计算机工程与设计》2008年第4期1019-1020,1024,共3页Computer Engineering and Design

摘　　要：金相图像处理是进行定量分析的前提,将金相图像中的目标物正确的提取出来对定量分析起着重要的作用。利用同态滤波器对亮度不均匀的铸铁金相图像进行滤波,校正因反射光强引起的亮度不均匀现象。基于数学形态学梯度算子提出新算法,利用预先定义的结构元素对图像中石墨边缘进行提取,之后经过后处理再进行填充。实验结果表明,该方法在保证石墨边缘的同时可以抑制噪声,能够有效提取出石墨的边界。Metallography image processing is the premise ofquantitative analysis. It is very important to pick up the object edge correctly. Using homeostasis filtering to filter the cast iron image, the influence of intensity uneven is decreased. Based on mathematical morphology gradient algorithm, a new method is proposed. It has been confirmed that this algorithm is applied to detecting the edge. At the same time, the influence of noise is reduced when the edge is examined.

关 键 词：图像处理 数学形态学 边缘检测 同态滤波 金相图像 

分 类 号：N911.73[自然科学总论]