基于GF-1遥感影像的艾比湖区域田间尺度土壤盐渍化监测方法

Soil Salinization Monitoring in the Ebinur Lake Region at A Field Scale Based on GF-1 Image

土壤盐渍化是制约农业生产和发展的主要障碍。目前土壤盐渍化的遥感监测主要基于中、低分辨率卫星影像,并采用传统的基于像元分类方法,对盐渍化信息的细节描述不足,监测精度不高。本文使用国产GF-1影像,结合自上而下的多尺度分割技术和面向对象的信息提取技术,针对田间尺度下的盐渍化信息进行精确地提取、分类,并与传统分类方法进行了对比。结果表明:面向对象法和最大似然法的分类总体精度分别为92.72%和84.31%,Kappa系数分别为0.90和0.78。该技术能准确提取田间尺度下的盐渍地信息,在未来的农田盐渍化高精度监测研究中具有一定应用价值和发展潜力。

Soil salinization is the main obstacle of agricultural production and development. At present, it is mainly based on remote sensing monitoring of soil salinization with low or middle resolution satellites, in addition to this, usual uses the traditional classification method based on pixels to monitor the salinization information, so, it is hard to get a detailed description and the monitoring precision is low. This paper use do- mestic GF-1 combining top-down multi-scale segmentation technology and object-oriented technology of in- formation extraction, aimed at the field scale accurately extract, salinization information classification, and compared with the traditional classification methods. Results show that the object-oriented method and maximum likelihood classification accuracy of 92. 72% and 84. 31% respectively, in general, the Kappa coefficient was 0. 90 and 0.78, respectively. The technology can accurately extract field scales salted information, in the future of farmland salinization high-precision monitoring has certain application value in the re- search and development potential.