基于知识的分层综合分类法在土地利用/土地覆盖遥感信息提取中的应用

土地利用/土地覆盖数据的获取是研究LUCC的重要基础工作。随着遥感技术的飞速发展,通过遥感提取土地利用/土地覆盖专题信息已成为LUCC研究必不可少的一步。目前遥感专题信息提取水平相对滞后于遥感数据获取,为了提高遥感数据在土地利用/土地覆盖的应用,寻找一种较好的、具有相对适用性的方法是目前遥感应用的一个迫切要求。本文比较了目前比较常用的几种土地利用/土地覆盖遥感信息提取方法,分别以西部干旱区(柴达木盆地)和东部地区(鄱阳湖地区)为例,提出在GIS支持下基于知识的分层综合分类方法,并通过和其他几种常用方法进行比较分析,得到如下结果:在自然环境相差较大的柴达木盆地和鄱阳湖地区,采用了GIS支持下基于知识的分层综合分类方法的提取精度均要比单独采用最大似然法、纹理分析法、神经网络分类法等方法的总体精度高出25%,Kappa系数高出0.2。由此可以说明了该方法对于土地利用/土地覆盖专题信息的提取是可行的,同时它也具有一定的适用性。

Application of knowledge-based multi-layer synthesis image classification in extracting remote sensing information of land use and land cover

The acquisition of Land Use and Land Cover (LUCC) data is a basic work to LUCC research.With the fast development of Remote Sensing (RS) technology, it's inevitable for LUCC research to acquire thematic information of LUCC through RS images.However, at present, RS thematic information's extracting fairly lags behind RS data acquisition.So, to improve the application of RS data in LUCC, a good and suitable way is urgent to find out.Taken the west arid area (Qaidam Basin) and the east humid area (Poyang Lake watershed) for example, several common RS information extracting approaches of LUCC are analyzed in this paper, then Knowledge-Based Multi-Layer Synthesis Image Classification Approach supported by GIS is put forward.It suggests that in Qaidam Basin and Poyang Lake watershed, whose environmental characteristics are great different, the extracting accuracy of Knowledge-Based Multi-Layer Synthesis Image Classification Approach is higher 25 per cent than that of Maximum Likelihood Classification, Texture Analysis and Neural Network Classification, and the coefficient of Kappa is higher 0.2.So, it proves that Knowledge-Based Multi-Layer Synthesis Image Classification Approach is feasible and suitable as it is used for extracting thematic information of LUCC in the research field.