一种改进的基于TM影像的水体提取方法

An Improved Method of Water Extraction Based on TM Image

针对现有基于TM影像的水体提取方法存在耗时、准确率低及通用性不强等问题,应用主成分分析(principal component analysis,PCA)法,并结合徐涵秋改进的归一化差异水体指数(modified normalized difference water index,MNDWI)构建了便于提取水体的假彩色合成图,通过分析水体在该合成图中的光谱特征,建立水体提取模型,从而实现基于TM影像的水体提取。比较改进方法与MNDWI、谱间关系(TM2+TM3)-(TM4+TM5)和新型水体指数等几种常见的水体提取方法分别在太原市城区、山区的水体提取效果,结果表明:改进方法更能快速、高效地提取TM影像水体,其精度高达95.57%。此外,该方法不需要对阈值进行人工选择,达到了高自动化的研究要求。

For time consuming,low accuracy and weak versatility of the existing methods of water extraction based on TM image, this paper combined principal component analysis （PCA） with the modified normalized difference water index （MNDWI） which was modified by Hanqiu Xu to construct a false color composite image which could separate water from others easily,realizing water extrac- tion based on TM image by analyzing the spectral characteristics of water in the false color image and establishing a water extraction model. By comparing the efficiency of this extracting water method with MNDWI, （ TM2 ＋ TM3 ） - （ TM4 ＋ TM5 ） and NWI which were applied in the urbar and mountain areas of Taiyuan respectively. The results showed that the proposed method could extract water body of TM image more rapidly and efficiently,with an accuracy up to 95.57%. In addition, this method did not require a manual selection threshold, reaching the research requirements of high automation.