基于多源信息空间的遥感影像自动解译研究

Study on automatic interpretation of remotely sensed imagery based on multi-information space

通过分析遥感影像自动解译的研究现状和存在的问题,提出了基于多源信息空间的遥感影像自动解译的研究方法,将遥感影像分为光谱信息空间、形状信息空间、区域地学信息空间、干扰信息空间等四个信息空间,通过对多源信息空间的数学描述和计算机处理实现了遥感影像的自动解译.

Extraction methods based on spectral character and single target figure character have been developed many years, but statistical method on classify is short of anticipation desire despite of getting some evolution. On the base of analyzing study actuality and question of automatic interpretation in remotely sensed imagery, it brings forward the method on automatic interpretation of remotely sensed imagery based on multi- information space which divides remotely sensed imagery into spectrum information space, shape information space, interfere information space and region geo- science information space. The study can realize automatic interpretation remotely sensed imagery if it can do description by mathematics and disposing by computer on multi - information space.