无人机多光谱成像仪图像的校正及配准算法研究  

An Algorithm of Image Error Correction and Registration for Unmanned Aerial Vehicle Multi-spectral Imager

作　　者：孙凡[1] 何志平[1] 戴方兴[1] 马艳华[1] 

机构地区：[1]中国科学院上海技术物理研究所,上海200083

出　　处：《红外技术》2006年第4期187-191,共5页Infrared Technology

基　　金：国家863"无人机海洋遥感监测技术"课题(2001AA636030)

摘　　要：利用光校参数和软件配准法对无人机多光谱成像仪图像进行校正和彩色合成。多光谱成像仪由3个光电头部组成,它们分别对同一地物的R、G、B谱段成像。基于无人机多光谱成像仪中各光谱段成像光电头部存在的性能差异,需要进行图像配准,才能保证图像合成的效果。具体分析与讨论光校参数法与软件配准法的特点,采用这两种方法进行图像配准,并在此基础上提出采用二者相结合的手段研究进一步的算法,以期达到更好的配准效果。The head of multi-spectral imager in the unmanned aerial vehicle is composed of three CCD cameras, which extract information of the same object in different spectral wavelengths. In order to enhance the image quality, the optical registration parameter method（ORPM） and soRware registration mcthod（SRM） were used to compensate the optical differences of CCD cameras. The characters of ORPM and SRM were discussed, and the two methods were combined in our study.

关 键 词：图像校正 配准 多光谱 几何标定 

分 类 号：TN29[电子电信—物理电子学]