多结构元多路加权形态边缘检测算法研究

Study of Edge Detection Algorithm Based on Multi-structure Element Multi-weighted Morphological

边缘检测是图像特征提取与模式识别的基础。形态学边缘检测因其算法简单,噪声抑制和边缘保持能力已成为研究热点。但很多算法采用单一结构元素,很难对弱边界进行有效处理。基于此,利用多结构元思想构造多结构元抗噪型数学形态学边缘检测器,很好地抑制了噪声和提取到弱边缘。

Edge detection is an essential foundation in image feature extraction and pattern recognition. Mathematical morphological edge detector is a heated researched field for its simple algorithm, dcnoising function and retaining ability for edge information. However, the operators are difficult to detect weak edge information because the general morphological methods adopt a single structure element. The paper uses multiple structuring elements thinking to construct a mathematic morphology edge detection operator with multiple structuring elements and noise elimination. The experiment results show that the detector effectively denoises and extracts weak edge information.