摘要
论述了基于红外图像背景边缘特征的块匹配配准算法.在图像背景中,利用灰度变化和边缘特征,并根据红外目标的亮度特征避开有目标的区域,搜索合适的图像块;在块匹配时,将图像块进一步细分,改进了相似度测量算法的准确性.实验证明,整个配准算法简单实用,准确有效.
A block matching registration algorithm based on background edge feature of infrared image is discussed. By the way of applying the gray-level changing, edge features and avoiding target region according to the intensity, the suitable image block in background can be searched. To make blocks match better, the algorithm improves the accuracy of similarity measurement through subdividing image block. The experimental results show that the registration algorithm is simple, practical, exact and effective.
出处
《光电技术应用》
2006年第2期51-54,共4页
Electro-Optic Technology Application
关键词
图像配准
边缘检测
块匹配
相似度测量
image registration
edge detect
block matching
similarity measurement
