基于全极化Radarsat-2数据的干旱区典型绿洲盐渍化信息提取

The Extraction of Saline Soil Information in Typical Oasis of Arid Area Using Fully Polarimetric Radarsat-2 data

土壤盐渍化不仅破坏了当地的自然生态环境,而且严重影响着可持续发展。为了对干旱区盐渍地进行有效、实时的监测,利用全极化合成孔径雷达PolSAR(Polarimetric Synthetic Aperture Radar)遥感数据进行盐渍地信息的提取。提出对以渭-库绿洲为研究区的全极化Radarsat-2数据进行Pauli分解和H/A/α分解,获取其极化特征分量。在此基础上,利用支持向量机(Support Vector Machine)、最小距离分类(Minimum Distance Classification)以及最大似然分类(Maximum Likelihood Classification)方法进行监督分类,并对分类结果进行对比分析。研究结果表明,极化分解方法在盐渍化监测中具有较好的研究潜力,基于Pauli和H/A/α分解的SVM分类方法更适用于PolSAR数据盐渍地信息的高精度提取。

Soil salinization not only does harm to the ecological environment,but also directly hampers the economic development.Fully polarimetric synthetic aperture radar ( PolSAR) remote sensing data was utilized in this research and salinized soil information in the typical oasis of arid area was extracted in order to carry out effective and real-time monitoring of salinized areas in arid zones.This paper presented and carried Pauli decomposition and H/A/α decomposition for the Radarsat-2 image data in the study area of Ugan-Kuqa river oasis,and the corresponding polarimetric parameters were acquired.On this foundation,supervised classification was carried out with support vector machine ( SVM) ,minimum distance classification and maximum likelihood classification,and then comparison and corresponding analysis were conducted among the classification results of all of them.The results of this research show that the polarimetric target decomposition methods proposed in this paper have some potential in monitoring of the soil salinization,SVM classification method based on Pauli and H/A/ α decomposition is more applicable for high-accuracy extraction of saline soil information by using fully PolSAR data.