摘要
图像配准技术是近年来发展迅速的图像处理技术之一,是图像融合、图像镶嵌、超分辨率图像处理等领域所不可缺少的关键步骤。对遥感图像自动配准领域出现的最新技术进行了总结,并将图像配准方法分为三类:基于区域灰度的配准方法、基于图像特征的配准方法和基于对图像的理解和解释的配准方法。从原理和配准性能上对各类方法进行了分析,并指出了它们各自的优势和存在的问题。
Image registration is one of the image processing technologies developed rapidly in recent years.It is an essential step in the fields of image fusion,image mosaic,and super-resolution image processing etc.The new automatic registration technologies for remote sensing images are summarized. Those technologies are divided into three categories：gray level-based regional registration,image feature-based registration,and image understanding and explanation-based registration.Their principles and registration performance are analyzed and their advantages and disadvantages are pointed out finally.
出处
《红外》
CAS
2010年第10期1-6,39,共7页
Infrared
关键词
图像配准
特征选择
算法
image registration
feature selection
arithmetic
