基于信息熵度量的高斯尺度空间角点检测算法

图像的多尺度表示是指从原始图像出发,导出一系列越来越平滑、简化的图像,这种简化意味着图像信息的丢失。图像角点是图像的一个重要的局部特征,本文结合尺度空间理论,实现了多尺度下Harris角点检测,并利用信息熵计算每一尺度下图像的信息量,信息熵是衡量不同尺度下图像信息的可靠方法。通过实验对提取出的特征点进行攻击,证明了这些点具有很好的鲁棒性,与单尺度Harris角点检测相比,该特征点具有更高的重复率,即尺度不变性。

Gassian scale-space corner detection algorithm based on info-entropy measurement

The basic idea behind a multi - scale representation is to embed the original image into such a oneparameter family of derived images, which become more and more smooth and simple. The simplification means the loss of details and information. Corner is an important local charactor to an image , combined with scale - space theory, this papper implemented multi - scale Harris oorner detection, and information under each scale was calculated through its entropy,it is a reliable method for scale- space. Attack was made to the detected feature points, and the experiment result shows these points have good robustness. Compared with single scale Harris corner detection, the points have better repeatability, which means they are scale invariant.