为了提高角点检测精度,增强算法对伪角点的抑制能力,在Noble算子的基础上充分考虑图像不同区域间灰度统计特性的差异,提出了一种基于局部熵和方差调整的Noble角点检测改进算法.该算法首先选择角点响应函数阈值以及非极大值抑制邻域大小的初始参考值,然后根据区域熵与区域方差分别估计这两个初始参考值的局部调整系数,最后用调整系数对随后的参考值进行加权,获得自适应于真实信号结构的局部阈值以及局部非极大值抑制邻域大小.实验结果表明,所提算法能够精确地检测出大部分真实角点,有效地消除伪角点的干扰.

In order to improve the accuracy of corner detection and enhance the ability of false corner suppression,an improved Noble corner detection algorithm based on image entropy and variance is proposed by taking into consideration the differences in statistical characteristics among different regions in gray images.First,the initial refe-rence values of the threshold of corner response function and of the window size for non-maximum suppression are respectively determined.Then,two fine-tuning coefficients corresponding to the two initial values are calculated for each region respectively according to its local entropy and variance.Finally,the reference values are weighted with the two fine-tuning coefficients,and the local threshold and its window size for non-maximum suppression,which adapt to signal structure,are obtained for each region.Experimental results show that the proposed algorithm can accurately locate most true corners and effectively eliminate the interference of false corners.