多角度成像仪MISR是美国对地观测计划中一个重要的传感器,它通过多角度观测提供了更多地表各向异性信息,其数据产品采用HDF-EOS文件格式,而且由于增加了角度维的数据,当前很多常用软件不能灵活地处理该数据,从而限制了该产品在我国的广泛使用。本文主要从应用角度出发,介绍了组件库和C语言结合的MISR卫星地面产品的处理方法,通过HDF-EOS库,可以灵活地处理MISR的高维数据;而对于MISR的坐标信患,可以通过GCTP库提供的坐标转换功能,方便地完成不同投影之间的转换。本文还通过高分辨率的SPOT分类图像作为先验知识,提取了MISR观测平面的不同地类的较纯像元的BRDF,结果表明,在MISR观测尺度上,植被的碗边效应和其他地类的一些信息有较好的表征。
The Multi-angle Imaging SpectroRadiometer (MISR) provides more 3-D information than single-angle sensors, and it was used widely now. The MISR products adopt HDF-EOS format, however, because of its multi dimension, many software can't process the data flexibly. In this paper, we introduced the processing method combined the COM component and C lan- guage. We used the HDF-EOS and GCTP library in the actual dataset and withdrowed the BRDF of pure pixels for 3 classes m the MISR observational plane. This research provides reference for the MISR product users in our country.