摘要
SUR算法是一种鲁棒性较好的特征提取算法,被广泛应用在图像处理和机器视觉上。但是,经典的SURF(SpeededUp Robust Features)算法积分图像(Integral Image)过程占用内存大而耗时多。为此,从积分图像处理方面对经典的SURF算法进行了改进,提出了只占用一个图像空间的快速积分图像算法。实验验证表明,该算法能够达到占用内存小且耗时少的预期效果。
SURF is a algorithm that can provide robust feature extraction and is widely used in image processing and machine vision. However, the classic SURF (Speeded Up Robust Features) algorithm Integral Image (Integral Image) process occupies more memory and time, therefore, in this paper, classic SURF algorithm is improved from integral image processing aspects and the fast Integral Image algorithm which only takes an Image space is put forward. Experiment results show that the algorithm can achieve expected effect of occupying less memory and time.
出处
《科学技术与工程》
北大核心
2013年第5期1350-1353,共4页
Science Technology and Engineering
基金
国家自然科学基金项目(61104119)资助
关键词
SURF算法
特征提取算法
积分图像
耗时
SURF algorithm
feature extraction algorithm
integral image
time-consuming