摘要
运动检测是目前研究的热门课题之一，已经提出了许多种检测方法，主要包括块匹配法、象素递归法和相位相关法。本文提出一种基于分割的傅立叶方法，该方法是根据傅立叶变换的基本性质在对图象进行分割的基础上来完成运动检测的，它能够分别处理物体平移和旋转，并能获得较准确的运动矢量。
The motion detection is the one of the hot topics of current research.Many detection methods have been presented which include the block matching algorithm,picture element recurrent and phase correlation algorithm.A new method has been given by the segmentation of Fourier transform,which can handle the displacement and rotation of the picture respectively,and this method can get more accurate motion vectors.
出处
《华南理工大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
1997年第9期62-66,共5页
Journal of South China University of Technology(Natural Science Edition)
关键词
傅立叶变换
块匹配法
象素递归法
相位相关法
Fourier transform
block matching algorithm
picture element recurrent
phase correlation algorithm
