基于多结构元素形态学的图像边缘检测算法  

Algorithm of image edge examination based on morphology multi-structural element

作　　者：韦宏强[1] 李彤华[2] 郑屹[2] 马宏[1] 

机构地区：[1]长春理工大学光电工程学院,吉林长春130022 [2]白城兵器试验中心2站,吉林白城137001

出　　处：《计算机工程与设计》2007年第22期5447-5448,5452,共3页Computer Engineering and Design

摘　　要：分析了出传统图像边缘检测算法的不足,利用形态学多结构元素的自然属性,自适应确定权重,然后采用腐蚀运算构造边缘检测算法,再将多结构元的检测信息加权求和。实验结果表明,与传统边缘检测算法相比,该算法在保持图像边缘清晰的同时,有很强的去除噪声能力,得到的结果边缘细腻、平滑,在细节和特征方面保持得更好。The theory and method of mathematical morphology is applied, natural quality of morphology structural element, self-adapt determines weight from the multi-structural element morphology transformation are used, then erosion operation is adopted to construct the edge examination algorithm, finally multi-structural element examination information is summed with weight. Experimental result indicates that self-adapt determined weight achieves better image processing effect than traditional method, has strong ability of eliminating noise as well as keeps clear image edge.

关 键 词：图像处理 数学形态学 边缘检测 结构元素 腐蚀 自适应权重 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]