YANG Li-jun,MA Ming-dong,TANG Li-jun.Research on Wetland Vegetation Classification of Chongming Eastern Tidal Flat Based on TM Image[J].Research of Soil and Water Conservation,2013,20(01):126-130.
基于TM影像的崇明东滩湿地植被分类研究
- Title:
- Research on Wetland Vegetation Classification of Chongming Eastern Tidal Flat Based on TM Image
- Keywords:
- vegetation classification; knowledge discovery; vegetation index; water index; neural network
- 分类号:
- TP751
- 摘要:
- 湿地植被遥感分类对于湿地生态环境的保护与管理具有重要意义，遥感湿地植被分类的难点是湿地环境复杂，植被光谱相似。以崇明东滩湿地为研究对象，利用FieldSpec 3 Hi-Res光谱仪和Landsat 5卫星影像，分析了湿地植被实测和遥感影像反射光谱曲线。在分析的基础上，采用基于知识发现的信息提取方法对湿地植被进行分类。首先将潮滩植被从遥感影像中提取出来；然后计算NDVI、DVI等植被指数，并进行典型植被可区分性植被指数评价；最后将最优植被指数（KT1，TRVI和DVI）作为辅助信息，对潮滩植被进行神经网络监督分类。研究结果显示，该方法的分类总精度较高达86.5%，具有一定的适用性。研究结果可为实现自动、半自动化植被分类与识别提供理论依据和技术支持。
- Abstract:
- Remote sensing classification of wetland vegetation is of great importance to wetland ecosystem conservation and management. The classification of wetland vegetation are difficult because of the intricate environment and the similar vegetation spectrum. Taking Chongming eastern tidal flat wetland as the study site, the spectral curves of on-the-spot survey and remote sense of wetland vegetation was analyzed by spectrometers and Landsat 5 Image. Based on the spectral curves, wetland vegetation was classified by methods of knowledge discovery-based information extracting. Firstly, information of wetland vegetation was extracted from remote images; then vegetation index such as NDVI and DVI was calculated and evaluated about the distinguishability of typical vegetation; and finally, taking the best vegetation index as auxiliary information, wetland vegetation was classified by neural network supervised method. The classification accuracy of this method was high, and it reached 86.5%, indicating that this method was universality, and can provide theoretical basis and technical way of automatic or semi-automatic vegetation classification and identification.
参考文献/References:
[1] 孟宪民.湿地与全球环境变化[J].地理科学,1999,19(5):385-391.
[2] 孙广友.中国湿地科学的进展与展望[J].地球科学进展,2000,15(6):666-669.
[3] 李晓兵,史培军.基于NOAA/AVHRR数据的中国主要植被类型NDVI变化规律研究[J].植物学报,1999,41(3):314-324.
[4] 刘瑜,韩震.基于遥感的长江口南汇潮滩植被群落时空动态变化[J].上海海洋大学学报,2009,18(5):579-585.
[5] 刘瑜,韩震,李睿.基于主成分分析和植被指数的潮滩植被信息提取研究[J].遥感应用,2010(4):44-50.
[6] Coops H, Hanganu J, Thdor M, et al. Classification of Danube Delta lakes based on aquatic vegetation and turbidity[J]. Hydrobiologia,1999,415:187-191.
[7] Dogan O K, Akyurek Z, Beklioglu M. Identification and mapping of submerged plants in a shallow lake using quickbird satellite data[J]. Environmental Magagement,2009,90(7):2138-2143.
[8] Lacapra V C, Melack J M, Gastil M, et al. Remote sensing of foliar chemistry of inundated rice with imaging spectrometry[J].Remote Sensing of Environment,1996,55(1):50-58.
[9] 张策,臧淑英,金竺,等.基于支持向量机的扎龙湿地遥感分类研究[J].湿地科学,2011,9(3):263-270.
[10] 阿也提古丽·斯迪克,赵书河,左平,等.基于知识工程师的崇明岛东滩自然保护区盐沼植被分类研究[J].国土资源遥感,2010(1):123-126.
[11] 刘瑜,韩震.植被指数在长江口潮滩湿地植被信息提取中的应用研究[J].遥感技术与应用,2009,24(6):777-783.
[12] 姚薇,李志军.Landsat卫星遥感影像的大气校正方法研究[J].大气科学学报,2011,34(2):251-256.
[13] 侯东,宋国宝,董燕生,等.TM传感器辐射定标参数精度分析[J].国土资源遥感,2010(4):14-18.
[14] 席晓燕,沈楠,李小娟.ETM+影像水体提取方法研究[J].计算机工程与应用,2009,30(4):993-996.
[15] 曹荣龙,李存军,刘良云,等.基于水体指数的密云水库面积提取及变化监测[J].测绘科学,2008,33(2):158-160.
[16] 李庆海,陶本藻.概率统计原理和在测量中的应用[M].北京:测绘出版社,1990.
[17] 李均力,盛永伟,骆剑承.喜马拉雅山地区冰湖信息的遥感自动化提取[J].遥感学报,2011,15(1):36-43.
[18] 田庆久,闵祥军.植被指数研究进展[J].地球科学进展,1998,13(4):327-333.
[19] Zhao Bin, Yan Yaner, Guo Haiqiang, et al. Monitoring rapid vegetation succession in estuarine wetland using time series MODIS-based indicators:an application in the Yangtze River Delta Area[J]. Ecological Indicators,2009,9(2):346-356.
[20] Aronoff S, Goodfellow W. Effective analysis of image processing to regional geochemical data[J]. Advanced Technology for Monitoring and Processing Global Environmental Data,1985,7:220-234.
[21] 张友水,冯学智,阮仁宗,等.Kohonen神经网络在遥感影像分类中的应用[J].遥感学报,2004,8(2):178-184.
[22] 张金屯,杨洪晓.自组织特征人工神经网络在庞泉沟自然保护区植物群落分类中的应用[J].生态学报,2007,27(3):1005-1010.
备注/Memo
收稿日期:2012-07-06;改回日期:2012-09-16。
基金项目:国家自然科学基金（41101359）；南京邮电大学项目（NY211073，2011XZZ11，2012XSG15）
作者简介:杨立君(1977-),男,黑龙江勃利人,博士研究生,讲师,主要研究方向:海岸带遥感技术应用。E-mail:ljyang@njupt.edu.cn