基于TM影像纹理与光谱特征和KNN方法估算5种红树林群落生物量
Biomass Estimation of Five Kinds of Mangrove Community with the KNN Method Based on the Spectral Information and Textural Features of TM Images
-
摘要: 为研究红树林生物量的遥感估算方法,本文提取广西和海南部分红树林TM遥感影像光谱及纹理特征,结合同地区地面调查的生物量数据,应用KNN方法,对生物量进行了遥感估算,并和多元逐步回归分析方法比较。研究表明:应用KNN方法估测精度随尺度的增大而增大,且K值取10优于K值取5;在像元尺度上,回归方法估测生物量优于KNN方法。Abstract: In order to research the biomass of mangrove based on remote sensing, the biomass of mangrove with the method of KNN was estimated by extracting spectral information and textural features from TM images, combining the field survey biomass data and compared with that of multiple regression analysis. The results showed that with KNN method, the accuracy increased with the extension of the scale, and K=10 was better than K=5 in the accuracy. Estimating biomass of mangrove in the pixel scale, the multiple regression analysis was better than that by using KNN method.
-
Key words:
- texture
- / KNN
- / biomass estimating
- / root-mean-square error
- / mean error
- / predication estimation accuracy
-
[1] 范航清.红树林一海岸环保卫士[M].广西:广西科技出版社,2000 [2] Rosenfeld A, Kak A. Digital Picture Processing (2nd edition)[M].Washington: Academic Press, 1982 [3] Lee J, Philpot W. Spectral textures pattern matching : A classifier for digital imagery[J]. IEEE Trans actions on Geosciences and Remote Sensing,1991,29:545-548 [4] Dengsheng Lu. Estimation of Forest Stand Parameters Using Landsat TM Images in the Brazilian Amazon Basin . International Symposium on Remote Sensing of Environment: Information for Risk Management and Sustainable Development; 2003-11-10-14; Honolulu, HI; US [5] 李明诗,谭 莹,潘 洁.结合光谱、纹理及地形特征的森林生物量建模研究[J].遥感信息, 2006(6):6-9 [6] Haralick R M. Statistical and Structural Approaches to Texture[J].Proceeding of the IEEE,1979,67:786-804 [7] Killki P,Paivinen R.Reference sample plots to combine field measurements and satellite data in forest inventory // SNS and Taksaattoriklubi, Remote Sensing-Aidec Forest Inventory. Hyytiala, Finland,1987:209-212 [8] Muinonen E, Tokola T. An application of remote sensing for communal forest inventory //SNS/IUFRO Workshop. The Usability of Remote Sensing for Forest Inventory and Planning, Umea, 1990:35-42 [9] Tokola T, Pitkanen J, Partien J, et al. Point accuracy of a non-parametric method in estimation of forest characteristics with different satellite materials[J]. J Rem Sensing,1996,17:2333-2351 [10] Isaaks E H, Srivastava R M. Applied Geostatistics[M]. London,Oxford University Press,1989 [11] Fazakas Z, Nillsson M, Olsson H. Regional forest biomass and wood volume estimation using satellite data and ancillary data[J]. Agriculture and Forest Meteorology,1999,98-99:417-425 [12] 温远光. 广西英罗港5种红树植物群落的生物量和生产力[J].广西科学,1999,6 (2):142-147 [13] 方红亮,张健挺,刘卫国.ERDAS遥感图像处理教程[M].北京:中国科学院地理研究所资源环境信息系统国家重点实验室,1998 [14] 徐新良,曹明奎. 森林生物量遥感估算与应用分析[J].地球信息科学,2006, 8(4):122-128 [15] 陈尔学,李增元,武红敢.基于K-NN和Landsat 数据的小面积统计单元森林蓄积估测方法[J].林业科学研究,2008,21(6):745-750 [16] Nilsson M.Simultaneous estimation of forest variables using Landsat TM data . Stockholm, Swedish University of Agriculutural Sciences, 1997
计量
- 文章访问数: 3551
- HTML全文浏览量: 293
- PDF下载量: 1595
- 被引次数: 0