摘要
针对SIFT(Scale-invariant featrue transform)多量性特征,可能会产生很多不必要的关键点,因此在本文中结合背景减法算法,对重点物体进行提取,更好的体现局部特征的优势。并且引入K-L变换,在不影响质量的情况下对特征向量进行降维处理,以提高运行效率。
The abundant characteristics of the SIFT（Scale-invariant featrue transform） may lead to many unnecessary keypoints.The background subtraction algorithm is presented to solve the problems.It extract the key objects,which better reflect the advantage of local property.At the same time,we introduce K-L transform,in the situation of without affecting the quality to carry on the reduced-dimension processing,to raises operating efficiency.
出处
《微计算机信息》
2010年第35期236-238,共3页
Control & Automation
