摘要
针对目前的SURF算法占用内存大和耗时长等缺点,提出一种基于兴趣区域的快速SURF算法。首先对目标图像进行阈值化分割算法进行分割,消除和抑制了图像中背景的无用信息,再利用SURF对分割后的图像进行匹配运算。
Since speed up robust feature （SURF） algorithm is applied in the process of feature matching and problems of wide range of searching,large volumes of data and slowly running occur,a new matching method of SIFT based on region of interest is pro-posed.In order to improve the performance for speed of algorithm,the image is segmented by threshold val-ue method which eliminates and suppresses useless in-formation of image background.Then,SURF is used to extract feature points of segmented image and match elementarily.
出处
《机械与电子》
2015年第3期39-43,共5页
Machinery & Electronics
基金
湖北文理学院大学生创新创业训练资助项目(201310519041)
关键词
SURF算法
阈值分割
图像匹配
实时性
speed up robust feature
thresholding
image matching
real time
