摘要
从成像光谱特性、成像分辨率和成像模式等方面对可见光、红外、高光谱和合成孔径雷达传感器的成像特点进行分析,根据一致性特征描述方法对多源遥感图像配准算法进行分类,指出多源遥感图像具有成像特性变化大、相关度小、匹配特征的空间分布不均匀等特点,其配准技术的关键在于提取不变的图像特征以及得到有效的匹配特征。
This paper analyzes the imaging characteristics of sensors including visible,infra-red,hyperspectral and Synthetic Aperture Radar(SAR) from aspects of imaging spectral properties,imaging resolution to imaging modality.The algorithms for multi-source remote sensing image registration are classified by consistent features.It gives conclusions that multi-source remote sensing images have properties of various imaging properties,low correlations and matching features distributed nonuniformity spatially.The key of multi-source remote sensing image registration lies in extracting stable image features and getting suitable matching features.
出处
《计算机工程》
CAS
CSCD
北大核心
2011年第19期17-21,25,共6页
Computer Engineering
基金
国家自然科学基金资助项目(40971245)
关键词
多源遥感图像
成像特点
一致性特征
图像配准
匹配特征
multi-source remote sensing image
imaging characteristic
consistency feature
image registration
matching feature