Inversion classifications of saline-alkali lands based on the hyperspectral data from the environmental mitigation Satellite

DOI

10.2991/icitmi-15.2015.202How to use a DOI?

Keywords

Remote sensing, regression prediction, the least squares support vector machine (SVM), saline-alkali land, Songliao basin, Daqing

Abstract

This study explores the best model for the quantitative inversion of soil salinity index in an area, and adapts the least squares support vector machines (LS - SVM) regression forecast method. A variety of index inversions is made in serious saline-alkali land in Daqing area, and decision-making method of binary tree is adopted to classify the saline-alkali by taking Songliao basin as an example. To have effective preventions and controls of saline-alkali land, based on hyperspectral data of the environmental mitigation Satellite( HJ - 1 A) we contrast the results of predictions between curve regression and least squares support vector machine (SVM) regression which are the two nonlinear regression modes in the prediction effects of salt content inversion. The results shows that the environmental mitigation satellite can be effectively used on the extraction about information aline-alkali land .The inversion model based on least squares support vector machine (SVM) is with highest accuracy; In RS, the method of decision binary tree can effectively be supported by using classifies saline-alkali lands. The results are accurate and reliable; Research shows that The salinization of Daqing area is serious, because the most parts of lands are alkalized soil, including mild alkaline lands , 345.03 km2 moderate alkaline lands ,1389.03 km2 and severe alkaline lands , 869.94 km2. The research has great significance for the prevention and cure of saline-alkali land and fast extraction information of saline-alkali land.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Song Wu
AU  - Qigang Jiang
PY  - 2015/10
DA  - 2015/10
TI  - Inversion classifications of saline-alkali lands based on the hyperspectral data from the environmental mitigation Satellite
BT  - Proceedings of the 4th International Conference on Information Technology and Management Innovation
PB  - Atlantis Press
SP  - 1195
EP  - 1202
SN  - 2352-538X
UR  - https://doi.org/10.2991/icitmi-15.2015.202
DO  - 10.2991/icitmi-15.2015.202
ID  - Wu2015/10
ER  -