线性迭代聚类和主成分分析的遥感影像变化检测  

Change detection for remote sensing images based on SLIC and PCA

作　　者：黄亮[1,2] 於雪琴 姚丙秀 杨泽楠 HUANG Liang;YU Xueqin;YAO Bingxiu;YANG Zenan(Faculty of Land Resource Engineering,Kunming University of Science and Technology,Kunming 650093,China;Surveying and Mapping Geo-Informatics Technology Research Center on Plateau Mountains of Yunnan Higher Education,Kunming 650093,China;Kunming Surveying and Mapping Institute,Kunming 650051,China)

机构地区：[1]昆明理工大学国土资源工程学院,昆明650093 [2]云南省高校高原山区空间信息测绘技术应用工程研究中心,昆明650093 [3]昆明市测绘研究院,昆明650051

出　　处：《测绘科学》2019年第11期189-194,共6页Science of Surveying and Mapping

基　　金：云南省应用基础研究计划面上项目(2018FB078);云南省教育厅科学研究基金项目(2016ZZX051);昆明理工大学引进人才科研启动基金项目(KKSY201521040);自然资源部地球观测与时空信息科学重点实验室经费资助项目(201911)

摘　　要：针对像素级变化检测方法对高空间分辨率遥感影像进行变化检测时存在检测结果过于细碎、错检和漏检过多等问题,该文提出一种结合简单线性迭代聚类(SLIC)和主成分分析(PCA)的对象级遥感影像变化检测方法。该方法首先采用SLIC算法分别对两期遥感影像进行分割得到超像素并标记颜色;然后采用PCA法对分割标记图进行降维并采用差值运算方法对两期分割标记图进行处理得到差异图;最后采用OTSU法对差异图进行二值分割得到最终的变化检测结果。为了验证实验结果,选取两组多时相遥感影像作为实验数据。实验结果表明,该方法得到的变化检测精度分别达到95.96%和91.9%,是一种可行的变化检测方法。Aiming at the problem that Pixel-level change detection method which was used to detect the change of high spatial resolution remote sensing images had many problems such as too small details,many false and missed result,an object-level change detection method for remote sensing images based on simple linear iterative clustering(SLIC)and principal components analysis(PCA)was proposed in this paper.First,the SLIC algorithm was used in this method to generate the super pixels and mark the color of the two temporal remote sensing images respectively;Then,the PCA algorithm was used to reduce the dimension of the segmented markup graph,and the difference map of the two temporal segmentation markings was obtained through difference method;Finally,the OTSU method was used to segment the difference map into two categories,and the final change results could be produced.In order to verify the experiment results,two sets of multi-temporal remote sensing images were selected as experimental data.The experimental results showed that the detection accuracy of the two sets of images change by using the proposed method could reach 95.96% and 91.9% respectively.So the proposed method was a feasible change detection method.

关 键 词：变化检测 遥感影像 主成分分析 简单线性迭代聚类 大津法 

分 类 号：P237[天文地球—摄影测量与遥感]