基于指数构建的高分辨率城市水体提取新方法

A New Method of High Resolution Urban Water Extraction Based on Index

针对高分辨率遥感影像水体指数适用性差异的问题,提出了一种水体指数CEWI和一种阴影指数MSRM,并综合2种指数构建了自动化城市水体提取模型NMEM。以国产资源三号卫星遥感数据为信息源,在对比分析各地类要素光谱信息的情况下,采用面向对象分割和特异性指数构建的方法,以北京、武汉、广州3市部分区域作为实验对象进行水体提取实验。通过与常用水体指数NDWI、高分辨率水体指数HRWI阈值提取以及依据光谱特性、几何特性等的随机树监督分类方法进行比较,实验结果表明,在各研究区内NMEM模型的水体提取效果最佳,用户精度均在91%以上,制图精度均在93%以上,Kappa系数均在0.98以上。该方法在准确提取水体的同时有效排除了大量黑暗阴影的干扰,基本消除了椒盐噪声。

Because of the limitation of previous water indices and the dark building shadow effect,it is a challenge to extract water bodies with high accuracy.The purpose of this study was therefore to devise a model that consistently improves water extraction accuracy in the presence of various sorts of environmental noise.Thus,we put forward a kind of water index(CEWI)and a shadow index(MSRM),and two comprehensive indexes to build an automation urban water extraction model(NMEM).Then,using the method of object-oriented segmentation and this model,parts of Beijing,Wuhan and Guangzhou were taken as the experimental area of water extraction.The performance was compared with that of the NDWI,HRWI and the random forest.The experimental results indicated that NMEM of water extraction is the best,with user accuracy above 91%,mapping accuracy above 93%,and Kappa coefficient more than 0.98.Experimental results show that the proposed method can increase extraction accuracy and reduce the influence of shadows.