收集了全球5个典型区域1980年、1990年、2013年的30m Landsat系列影像数据集，对其进行了几何纠正、图像增强等基本处理；收集了典型区域的野外考察资料和参考数据；利用马尔科夫随机场融合空间上下文信息、降尺度NDVI时序数据生成技术发展了一套基于遥感信息变化自动检测与一级类更新技术；基于多源遥感数据的光谱、时相和纹理信息以及生态、气候地理分区等辅助信息，发展多源遥感信息与地学知识复合应用的地表覆盖二级类信息分类技术方法；面向地表覆盖精细化制图的网络化检核和验证需求，研究了异质数据的服务化集成技术。%Collected the five typical regions in 1980, 1990, 2013 of 30 m series of Landsat image data sets, and finished the basic processing such as geometric correction、 image enhancement; collected the field survey data and reference data of typical area; used markov random field space context information fusion and the downscaling NDVI time-series data generation technology developed a set of automatic change detection and A-level class update based on remote sensing information; based on multi-source remote sensing data of spectrum, and the texture information and ecology, climate, geographical division and other auxiliary information, developed surface coverage sub-type class information classification method with combining geological knowledge with multi-source remote sensing information; to satisfy surface coverage fine mapping network check and validation requirements, developed the heterogeneous data integration technology of service.
展开▼
