摘要
对红外、可见光两种不同谱段图像的匹配技术进行研究。在比较不同预处理方法后,采用边缘增强而不是边缘提取的预处理技术,以增强图像的有效边缘信息,突出不同谱段图像的共性特征,然后使用去均值归一化互相关的办法进行相似性度量。在实验中,通过确定影响匹配正确率的关键指标,对不同预处理方法和匹配策略得到的匹配结果进行比较分析,验证方法的有效性。
A scene matching technique by edge enhancement for infrared image and optical image is described in this paper. Compared with different image pre - processing methods, an edge enhancement approach is adopted in order to increase valid edge information and intensify the common characteristics between infrared image and optical image. And then, cross correlation is used as similarity measures for scene matching. With some criteria for the matching probability, the performance of different pre - processing methods and different matching strategy is evaluated. The experimental results show the validity of the proposed approach.
出处
《计算机与数字工程》
2006年第12期30-32,共3页
Computer & Digital Engineering
基金
国防基础研究项目(编号A1420061266)资助
关键词
景象匹配
边缘增强
多谱图像
scene matching,edge enhancement,multi - sensor image
