基于遥感影像分割单元的土地利用变化快速检测方法  

Rapid detection method for land use change based on remote sensing images segmentation units

作　　者：王志杰[1] 

机构地区：[1]南京林业大学土木工程学院,江苏南京210037

出　　处：《南京林业大学学报（自然科学版）》2015年第3期1-5,共5页Journal of Nanjing Forestry University：Natural Sciences Edition

基　　金：国土资源公益性行业科研专项项目(201211028-6)

摘　　要：针对土地利用变化动态监测的实际业务需求,提出一种土地利用变化快速检测方法,即基于同一区域土地利用现状图与高分辨率遥感影像进行叠加分析,通过土地利用现状图图斑边界对高分辨率遥感影像进行分割,并对分割单元的像素灰度值进行统计分析,以此判断图斑内对应的土地利用是否发生变化,从而提取影像发生变化区域范围。以某城市局部区域土地利用变化监测为例,通过对该区域2013年高精度遥感影像与2012年土地利用现状图叠加分析,对938个图斑进行土地利用变化检测,发现变化图斑12个。其中,8个为实际发生变化图斑,4个为错判,无漏判。实验结果表明:该方法能快速、有效地实现土地利用变化信息的检测和提取,并适用于识别提取新增建设用地占用农林用地的情况。Because of actual demand of land use change supervision, a method for quick detection of land use change is presented. This is the analysis of the remotely sensed imagery that is overlapped with corresponding land-use map. The image is segmented by pattern spots in the presented land-use map, and the gray .level of pixel of every segment unit is obtained by statistics and analysis. Then, the land use of pattern spot is judged whether it has changed. Finally, the changed pattern spots in remotely sensed imagery are extracted. The high-resolution remotely sensed imagery of 2013 and present land-use map of 2012 from a certain county are selected as an example to extract the section. Results showed that among 938 experiment pattern spots, 12 pattern spots are judged to be changed, among 12 pattern spots, eight of them are changed actually, four of them are not changed actually, and no actually changed pattern spots are ignored. The ex- periment showed that the method could quickly and effectively detect and extract the information of land use change, especially was suit for the situation of farmland which was changed to be construction land.

关 键 词：土地利用变化 高分辨率遥感影像 变化检测 统计分析 

分 类 号：F301.2[经济管理—产业经济] P208[天文地球—地图制图学与地理信息工程]