首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于相似像素点的Harris角点检测改进算法
引用本文:朱聪.基于相似像素点的Harris角点检测改进算法[J].大众科技,2014(3):28-31.
作者姓名:朱聪
作者单位:同济大学电子与信息工程学院,上海201804
摘    要:Harris角点检测需要计算图像中每一个像素点的角点响应函数值(CRF),这使得算法运行速度慢,不能满足实时性的要求。针对这个缺点,文章提出了一种改进算法。改进算法通过分析图像中每个像素点8邻域范围内相似像素点分布情况,剔除那些非角点的像素点,并选出潜在的角点像素点作为下一步Harris检测的候选点,从而大大提高了算法的运行效率。

关 键 词:角点检测  图像拼接  相似像素点  Harris

An improved algorithm of Harris corner detection based on similar pixel
Abstract:Harris corner detection algorithm needs to calculate corner response function of each pixel, which makes the algorithm low efficiency and poor real-time performance. An improved algorithm is proposed to overcome this shortage. The improved algorithm is used to analyze distribution of the similar pixel which gray level is similar to the target pixel within 8 neighborhoods of each pixel, then the candidate corner pixels are selected for following steps of Harris corner detection, thus the efficiency is highly improved.
Keywords:Harris  Corner detection  image stitching  similar pixel  Harris
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号