摘要
基于微分算子的边缘检测方法存在抗噪性能弱等缺点,非线性的数学形态学边缘检测可以克服这些缺点,而使用单一结构元素对图像进行形态学处理会模糊很多细节。鉴于这些不足,运用多结构多尺度的思想,将形态学的滤波和边缘检测结合起来,提出一种边缘检测方法,首先采用多结构多尺度的方法对噪声图像进行串联开闭滤波,再利用改进的多结构元素的形态学梯度算子进行边缘提取。实验结果表明,该方法具有较好的抗噪效果且提取的边缘比较平滑。
The method of edge detection based on differential operators has disadvantages such as weak anti-noise,and the edge detection based on nonlinear mathematical morphology can overcome these shortcomings,but morphological image processing only with a single structure element will make many of the details blurred.In view of the weakness,using an idea of multi-structure and multi-sale,combining morphologic filter with edge detection,a novel method of edge detection is proposed.First,the noise can be filtered by the series connection morphological open-close filter with a multi-structure and multi-scale approach,and then an improved morphological gradient operator based on multi-structure element is used for edge extraction.The experimental results showed that the proposed method has better anti-noise effect and the edge extracted is quite smooth.
出处
《计算机与数字工程》
2012年第2期102-104,共3页
Computer & Digital Engineering
关键词
边缘检测
形态学
多结构
多尺度
edge detection
morphology
multi-structure
multi-scale
