摘要： 现有Harris特征点检测算法采用高斯滤波进行平滑，图像存在角点信息丢失与偏移的现象。为解决该问题，提出基于变分B样条滤波与快速局部窗口搜索相结合的Harris特征点检测算法，选择具有低通特性的B样条函数作为平滑函数构造滤波器，引入形态学滤波中的极大值滤波思想，利用快速局部窗口搜索算法进行特征点局部极值的提取，从而提高特征点提取的精度和速度。实验结果表明，改进算法具有特征点提取快速均匀、检测定位准确、抑噪性好的特点。

关键词: 变分B样条函数, Harris算法, 特征点检测, 局部极值

Abstract: Existing Harris feature point detection algorithm uses Gaussian smoothing filter, but it has a phenomena that image feature points extracted angular point information is of lost and migration. In order to solve the problem, the data-fitting and low-pass characteristics of B-spline function are introduced into the algorithm to construct a new filter, while the morphological filter is also introduced into the maximum inhibition of ideal that using fast moving window search algorithm for extracting feature points to improve the precision and speed of features point extraction. Experimental result shows that the improved algorithm detected feature points is rapid, uniform and accurate, and it has good noise suppression.

Key words: variational B-spline function, Harris algorithm, feature point detection, local extreme value

中图分类号: 

• TP39