Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Chine → 2007 → The Application Study On Extraction Of Saline Soil Information In The Arid Region Using Satellite SAR Images

Xinjiang University (2007)

The Application Study On Extraction Of Saline Soil Information In The Arid Region Using Satellite SAR Images

何祺胜; He Zuo Sheng

Titre : The Application Study On Extraction Of Saline Soil Information In The Arid Region Using Satellite SAR Images

Auteur : 何祺胜; He Zuo Sheng

Grade : Master’s Theses 2007

Université : Xinjiang University

Résumé partiel
The soil salinization is the land degeneration which is caused by the unreasonable human activity and the frail ecological environment in the arid and semi-arid area, which is nowadays as an important environmental and social economy question as the desert in the world. According to the statistics,there are 952 million km2 salinization land in the whole world,and China is one of countries which are affected seriously, where the question of the soil salinization is still extremely prominent. The soil salinization is comprehensively concerned by more and more people in the whole world because it brings the destructive effect to the social economy, environment and the agriculture ecosystem. It is very important to father saline soils, prevent further pervasion and degeneration, continue to develop the agriculture, and carry on the reasonable plan using remote sensing high-tech technology, which is characteristic by its macroscopic, comprehensive, dynamic and fast quality, to study the quality, the scope, the area, the geographic distribution, the dynamic change and the degree of saline soils.This article uses the radar images as the most important data, which is also the application of radar images for the first time in arid area of Xinjiang. This article also comprehensively uses the superiority of other remote sensing and the spatial data, enhances the extraction precision of saline soils information, thus comprehensively explores and evaluate quantitatively the soil salinization in the arid and semi-arid area, develops applied research on the extraction of saline soils information using microwave remote sensing, provides the highly effective decision system ,which serves for the sustainable development, to all levels of governments and the policy establishing department, and makes the contribution for the western big development.1. Analyze the elimination of SAR images spot noise with different filter method and different window size. In order to eliminate the SAR image spot noise, using the Lee filter, the enhanced Lee filter, the Frost filter, the enhanced Frost filter, the Gamma filter, the Kuan filter, the Local Sigma filter, the Bit Errors filter to process the radar image, this article has analyzed the S.D change under different window sizes, then taken the indexes of FI and ESI under different window sizes to analyze the change of the typical things of severe saline soil, middle and light saline soil, farmland, water body and the clay. The result indicated that the effect of the enhanced Frost filter and the enhanced Lee filter to remove the spot noise was satisfied.2. Analyze the application of decision tree classifier on the extraction of saline soils based on multiple source data in arid area. Firstly salinization causes were discussed. The results shows that arid desert climate is a prerequisite to the formation of salinization, parent rock containing salt is the material basis for the formation of salinization, surface water and groundwater supplies are the impetus to the formation of salinization, human factors is a key factor to the formation of secondary salinity in irrigation areas. Through contrast on whether the radar waveband was used on decision tree classifier the result indicated that after the radar waveband was joined to carry on principal component transformation, the fifth principal components can separate the water body information and the severe saline soil information, and can extract the severe saline soil information effectively. Through the decision tree classifier model established, the model which was joined by the radar waveband has less pitch points, reduces the operation process, makes the entire extraction flow of saline soils information more concise, and the selection of characteristic waveband was also more convenient and the precision of extraction also has remarkable increase.3. Analyze the application of multiple source information fusion on the extraction of saline soils based on in arid area. Firstly the paper carries on the outline to the commonly used remote sensing image fusion method, then uses the fusion method of HIS, Brovey and PCA to fuse the radar image and optics image to enhance the precision of extraction of saline soils. The result indicated that the picture based on the principal component fusion had enhanced the resolving power greatly, and enhanced the discriminations between the saline soils and other things.

Mots clés : saline soils information extraction; radar image processing; multiple source data fusion; decision tree classifier; ion response;

Présentation (CNKI)

Page publiée le 22 avril 2013, mise à jour le 21 octobre 2017