Remote Sensing Image Classification Based on SVM Method for Ecological Service Forests

doi:10.13466/j.cnki.lyzygl.2009.01.020

Abstract

Abstract: This paper deals with the RS image classification based on the SVM method, using space characteristic information for classification of IKONOS high spatial resolution images and monitoring of public w elfare forests.Analy sis was conducted on comparison of this method with tradition method. The resultshow s that the RS image classification based on the SVM method can solve the image classification fragmentation, low accuracy etc, and has advantage in study speed, orientation abili ty and expression, etc.The aim of this paper is to discuss a method to inquire into the classification method of public welfare forests with high spatial resolution RS image and providing theoretical basis and data support for the development of forestry information netw ork and “digi tal forestry”.

Key words: forest ecosystem, ecological service forests, SVM, space characteristic, superior super flat surface

CLC Number:

TP79

REN Qiong, JIANG Hong, CHEN Jian, LI Tusheng, PENG Shikui, YU Shuquan. Remote Sensing Image Classification Based on SVM Method for Ecological Service Forests[J]. FOREST RESOURCES WANAGEMENT, 2009, (1): 107-113.

References

[1] 周国逸, 阎俊华.生态公益林补偿理论与实践[M].北京:气象出版社, 2000:5-6.
[2] 张正寿, 刘安兴.浙江林业自然资源(森林卷)[M].北京:中国农业科学技术出版社, 2002:9-31.
[3] Zhang Y.Opt imisation of building det ection satellite images by combining mult ispectral classification and texture filtering [J]. PE & RS, 1999, 54:50-60.
[4] 黄慧萍, 吴炳方, 李苗苗.高分辨率影像城市绿地快速提取技术与应用[J].遥感学报, 2004, 8(1):68-74.
[5] Vapnik V N.The nature of statist ical learning theory[M].New York:Springer-Verlag, 1995.
[6] Burges C J C.A tutorial on support vector machines for patt ern recognition[J].Know ledge Discovery and Dat a Mining, 1998, 2 (2).
[7] Cortes C, Vapnik V.Support vector networks[J].Machine Learning, 1995;20(3):273-297.
[8] Scholkopf B, Plat t J C, Shawe Taylor J, et al.Estimating the support of a high-dimensional distribution[J].Neural Computation. 2001, 13(7):1443-1471.
[9] Seidel K, Schroder M, Schwarz G, et al.Query by image content from remot e sensing archives[J].Proc IEEE IGARSS, 1997, 97 (1):616-618.
[10] Manjunath B S, Ma W Y.Browing large satellit e and aerial phot ographys[J].IEEE Int Conf on.Image Processing, 1996, 2:765-768.
[11] 刘江华, 程君实, 陈佳品.支持向量机训练算法综述[J].信息与控制, 2002, 31 (1):45-50.
[12] 张锦水, 何春阳, 潘耀忠, 等.基于SVM 的多源信息复合的高空间分辨率遥感数据分类研究[J].遥感学报.2006(1):49- 57.
[13] S trang G.Introduction to applied mathemati cs[M].Wellesley, MA:Wellesley-Cambridge Press, 1986.
[14] 刘江华, 程君实, 陈佳品.支持向量机训练算法综述[J].信息与控制, 2002, 31(1):45-50.
[15] Gong P, How arth P J.The use of structural information for improving land-cover classification accu racies at the ruralurban f ringe[J].1990, 56(1):67-731.
[16] Chang C C, Lin C J.LIBSVM:a libray for Support Vector Machines[EB/OL].http://www.csie.ntu.edu.tw/～cjlin/libsvm/.
[2008-09-28].2004.
[17] Kerrthi S S, Lin C J.Asym ptotic behaviors of Support Vector Machines with Gaussian Kemel[J].Neural Computation, 2003, 15 (7):1667-1689.
[18] Joachims T.Making Large-Scale SVM Learning Pract ical [C]//.Scholkoph B, Burges C, Smola A Advances in Kermal Methods Support Vector Learning.Botson Ma Mit Press, 1999.
[19] Jiang H, Liu S R, Sun P S, et al.The influence of vegetation type on the hydrological process in landscape scale[J].Canadian Journal of Remote Sensing 2004, 30:743-763.
[20] Jiang H, Srittholt J R, Frost P A, et al.The classi fication of lat eserial forests in the Pacific northwest, USA using Landsat ETM + imagery[J].Remote Sensing of Environment, 2004, 91:320- 331.
[21] Kwang I, Keechul J, Hang J.Support Vector Machines for texture classification[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, 24(11):256-264.