DU Xian-rong1,2
(1. School of Information and Communication Engineering, North University of China, Taiyuan 030051, China; 2. Optoelectronic Information and Instrument Engineering Technology Research Center in Shanxi Province, Taiyuan 030051, China)
Abstract： Aiming at the problems of high time-consuming, low accuracy and weak versatility of the existing methods of water extraction based on TM image, this paper combines principal component analysis (PCA) with the modified normalized difference water index (MNDWI) which was improved by XU Han-qiu to construct a false color composite image that could separate water from others easily. This method can realize the water extraction based on TM image by analyzing the spectral characteristics of water in this false color image and establishing a water extraction model. This paper also compares the efficiency of this method with MNDWI, (TM2+TM3)-(TM4+TM5) and new water index (NWI), which were applied in the city and mountain of Taiyuan, respectively. The results show that the proposed method can extract water body from TM image more rapidly and efficiently and its accuracy is up to 94.03%. In addition, this method does not require a manual selection threshold, which meets the research requirement of high automation.
Key words： TM image； water extraction； principal component analysis (PCA)； modified normalized difference water index (MNDWI)
CLD number： TP751 Document code： A
Article ID： 1674-8042(2014)03-0048-05 doi： 10.3969/j.issn.1674-8042.2014.03.009
References
[1] DENG Kun. Remote sensing image classification based on multi-source data. Northwest Agriculture and Forestry University in Shanxi Province, 2008.
[2] BI Hai-yun, WANG Si-yuan, ZENG Jiang-yuan, et al. Comparison and analysis of several common water extraction methods based on TM Image. Remote Sensing Information, 2012, 27(5): 77-82.
[3] YANG Cun-jian, XU Mei. Discussion on water extraction based on remote sensing information mechanism. Geographical Research, 1998, 17(Suppl): 86-89.
[4] ZHONG Chun-qi. Study of water extraction in wetland automatically from TM images. Water Resources Research, 2007, 28(4): 589-595.
[5] WANG Zhi-hui, YI Shan-zhen. Comparison and research on the different index models used in water extraction by remote sensing. Science Technoligy and Engineering, 2007, 7(4): 534-537.
[6] OU Chun-jian. Research on principal component analysis of remote sensing images processing through example. Geomatics and Spatial Information Technology, 2006, 29(5): 56-59.
[7] XU Han-qiu. A study on information extraction of water body with the modified normalized difference water index (MNDWI). Journal of Remote Sensing, 2005, 9(5): 589-595.
[8] ZHONG Jia-qiang, WANG Run-sheng. Multitemporal remote sensing images change detection based on ICA. Journal of Electronics and Information Technology, 2006,28(6): 994-998.
[9] JIA Ke-li, CHANG Qing-rui. Extracting water body informa-tion using principal components analysis. Journal of Huaiyin Teachers College (natural science editon), 2006, 5(3): 242-245.
[10] Mcfeeters S K. The use of the normalized difference water index(NDWI) in the delineation of open water feature. International Journal of Remote Sensing, 1996, 17(7): 1425-1432.
[11] DING Feng. A new method for fast information extraction of water bodies using remotely sensed data. Remote Sensing Technology and Application, 2009, 24(2): 167-171.
TM影像水体的提取研究
杜先荣1,2
(1. 中北大学 信息与通信工程学院， 山西 太原 030051;2. 山西省光电信息与仪器工程技术研究中心， 山西 太原 030051)
摘要: 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， 应用主成分分析法（principal component analysis， PCA）， 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。 比较本文方法与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： 本文提出的方法更能快速、 高效地提取TM影像水体， 其精度高达94.03%。 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。
关键词: TM 影像； 水体提取； 主成分分析法(PCA)； 改进的归一化差异水体指数(MNDWI)
引用格式：DU Xian-rong. Water bodies extraction from TM images. Journal of Measurement Science and Instrumentation, 2014, 5(3): 48-52. [doi: 10.3969/j.issn.1674-8042.2014.03.009]
[full text view]
