摘要
传统的边缘检测算法存在检测边缘不连续、漏检、抗噪性能差等缺点。针对传统方法在边缘检测中遇到的问题,结合数学形态学的方法,介绍一种基于多尺度灰值形态学的边缘检测算子,该算子是在形态学梯度算子的基础上改进之后,结合多尺度的概念,将各种不同尺寸下的边缘图像结合起来,提取出比较理想的边缘图像。采用Visual C++编程实现算法,并与传统边缘检测算子进行比较。实验结果表明本文提到的算子检测边缘效果更好。
The traditional edge detection method exist some shortcomings,such as： undetected edge,discontinuous edge detection and poor anti-noise performance.According to these problems in edge detection,combined with mathematical morphology method,this paper introduces a kind of the edge detection operators based on multi-scale grayscale morphology.On the basis of emended morphological gradient,combined with multi-scale,the operator will combine the different size edge image,extract the ideal edge image.This paper adopts visual c＋＋ programming algorithm,and compares with the traditional edge detection operators.The experimental results show that this edge detection operator works better.
出处
《山西电子技术》
2011年第5期25-26,45,共3页
Shanxi Electronic Technology
关键词
边缘检测
灰值形态学
多尺度
形态学梯度
edge detection
grayscale morphology
multi-scale
morphological gradient
