基于遥感影像多尺度分析技术的LUCC研究

基于遥感和地理信息系统技术研究土地利用／覆被变化（LUCC）目前普遍采用的方法，其中研究区不同时期遥感影像土地利用／覆被信息的获取是研究的关键，而传统的单一尺度的信息自动提取方式很难满足高精度LUCC研究的需要。为此，提出了以多尺度分析技术来提取土地利用／覆被信息的方法，并以江津市的LUCC研究为例，显示出这种方法的优越性。

Multi-scale Analysis of Remote Sensing Data for LUCC Detection

Most studies of LUCC(Land Use and Cover Change) are based on remote sensing and GIS technologies at present.Withdrawing information from the remote sensing graphical data is very important for they relate to the cost and the objective reliability of results.And it is very difficult to withdraw the information automatically using traditional single scale analysis technique to meet the need of high accuracy LUCC study.For this reason,a new method named multi scale analysis technique was brought forward to get land use and cover information.Graphical analysis with single scale could not meet the heterogeneity and objects extraction.To acquire them for the LUCC research multi scale image analysis is necessary.It creates meaningful objects and builds a hierarchy levels close to land cover character using multi scale network analysis model.Objects could be analyzed in their suitable scale levels.The potential advantage of this method is obvious that classification is based on multi scale and the meaningful combination objects,rather on the pixels.In multi scale graphical analysis,the size of "objects" is not necessarily the same as the pixels,but altered by the combination scale.As an example to make the application to practice,in the latter half of this paper we use it to study the LUCC of Jiangjin city among the ten years between 1992 and 2002,proved its superiority.