摘要
基于图像信息的目标跟踪问题中,对目标特征的精确提取与识别是极为重要的.本文针对点状特征的提取,采用了一种分阶段的高精度亚像素特征点提取方法.具体做法是:首先,利用Harris算子,获得像素级特征点;然后利用像素级特征点附近亮度值的梯度信息,进行亚像素级特征提取.实验结果表明,这种分阶段实现的亚像素级特征提取方法可以保证提取到尽可能多且位置精确的的亚像素级特征点.
Aiming at the need of point feature extraction, a method of extracting highly accurate subpixel feature point by stages was adopted. Firstly, pixel feature points were obtained by applying Harris arithmetic;And then subpixel feature points can be picked up by using the luminance grade information near these pixel feature points. All outcomes of experiments show that precise and quite exact subpixel feature points can be assuringly obtained by using the method put forward in this paper.
出处
《南华大学学报（自然科学版）》
2006年第2期54-57,共4页
Journal of University of South China：Science and Technology
关键词
图像处理
特征提取
亚像素
image processing
feature extracting
subpixel
