Harris算法和Susan算法的实现及分析

Implementation and Analysis of Harris Algorithm and Susan Algorithm

角点检测在机器视觉和图像处理的许多方面都起着十分重要的作用。论文主要研究了数字图像的角点特征检测方法,应用了Matlab软件对图像进行处理。通过应用Matlab编写程序,论文编辑有Harris算法和Susan算法的程序。通过Matlab运行得出图像提取角点的结果,分析了两种算法的优缺点。Harris角点检测原理是对于一幅图像,角点与自相关函数的曲率特性有关,自相关函数描述了局部图像灰度的变化程度。Susan算子是一种基于灰度的特征点获取方法,适用于图像中边缘和角点的检测,可以去除图像中的噪声,它具有简单、有效、抗噪声能力强、计算速度快的特点。

Corners are image points that show a strong intensity change.This paper mainly studies the method of digital image feature detection and processed image by Matlab.By means of the use of Matlab,this article editing program had Harries algorithm and Susan algorithm.The results of image corners are obtained by Matlab,and the advantages and disadvantages of the two algorithms are analyzed.The Harris corner detection principle is related to the curvature characteristics of the angular point and auto-correlation function.The auto-correlation function describes the degree of variation of the local image gray degree.Susan operator is a kind of method based on gray level for acquiring characteristic points,applying to the image edge and corner detection,which can eliminate the noise in the image,it has the characteristics of simple,effective,anti-noise ability and fast calculation.