光谱模型结合面向对象法的山区水体提取  

Water Extraction in Mountainous Area Based on Spectral Model and Object-oriented Method

作　　者：董哲 王凌[1] 朱西存[1] 冯文斌 张美霞 DONG Zhe;WANG Ling;ZHU Xichun;FENG Wenbin;ZHANG Meixia(College of Resources and Environment,Shandong Agricultural University,Taian,Shandong 271000,China)

机构地区：[1]山东农业大学资源与环境学院,山东泰安271000

出　　处：《遥感信息》2022年第4期121-127,共7页Remote Sensing Information

基　　金：国家自然科学基金面上项目(42171378)。

摘　　要：山区因地势起伏大、水体分布零散导致遥感提取水体信息精度不高。另外,对于高分二号(GF-2)影像,受限于只有4个波段,无法构建已有水体提取精度较高的指数。鉴于此,以泰山为研究区,采用GF-2影像,提出差异水体光谱模型结合面向对象法的水体信息提取方法,并与阴影水体指数决策树、改进的阴影水体指数决策树以及支持向量机3种方法进行对比。结果表明,该方法能够有效去除山体阴影的影响,较好地保持了水体信息,对细微水体也有良好的提取能力,在实验和验证影像中总体精度分别达到98.02%和97.33%,Kappa值分别达到0.9533和0.9334,均高于其他3种方法。该方法在准确提取水体的同时,有效减少“椒盐现象”的发生,可为类似山区水体提取提供一定的参考。Due to the large relief and scattered water in mountainous areas,the extraction accuracy of water information is not high.Limited to only four bands,GF-2 images are unable to be used for a more accurate existing water index.To solve the problems,a difference water spectral model combined with object-oriented method is proposed to extract water on Mount Tai based on GF-2 images,and three methods,shadow water index decision tree,improved shadow water index decision tree and support vector machine,are used for comparison.The results show that the method could effectively remove the influence of mountain shadow and keep the water information well,especially has good extraction ability for tiny water.The overall accuracy of the test and validation images is 98.02%and 97.33%,respectively,and Kappa is 0.9533 and 0.9334,respectively,which is higher than those of the other three methods.This method can not only accurately extract the water,but also effectively reduce the occurrence of salt-phenomenon,which could provide a certain reference for the extraction of water in similar mountainous areas.

关 键 词：GF-2影像 水体提取 差异水体光谱模型 面向对象法 泰山 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]