Research On Karst Rocky Desertification Information Detection With Remote Sensing

Karst rocky desertification is one of the main types of land desertification, and it is one of the main ecological problems in Karst area of Southwest China. Not only restricts the development of local social economy, but also seriously threatens people's living environment. Using remote sensing technology to analyze the situation of rocky desertification in Karst has become an effective technical method, which provides a scientific basis for the management and prevention of rocky desertification in Karst area. In this paper, the remote sensing image of the Guilin, which is provided by the Landsat5TM satellite in the United States, is used as the data source, using the remote sensing technology and the computer technology to extract and analyze the information of rocky desertification in this area. On multi feature of karst features of satellite images using BP neural network was selected for the extraction of the karst region. Experimental results show that in complex spatial distribution of karst area, according to the remote sensing image based on BP neural network of karst mountains classification, the classification accuracy of multi feature combination to higher accuracy than single feature combination. Since there is no unified rocky desertification evaluation index and rocky desertification classification index, so according to previous experience and rocky desertification evaluation index selection principle, the study area of karst rocky desertification divided into five grades:no rocky desertification, light rocky desertification, moderate desertification, rocky desertification strength, strong rocky desertification. At the same time, the use Dimidiate Pixel Model of vegetation cover and bare rock ratio as evaluation index of rocky desertification. The study area is obtained according to the two evaluation index of Karst rocky desertification distribution. Analyze the karst normalized difference vegetation index (NDVI), improved enhanced vegetation index (GEVI), modified soil adjusted index (MSAVI), difference vegetation index (DVI), green vegetation index (GVI), ratio vegetation index (RVI), soil brightness index (BI), normalized rock index (NDRI), vegetation coverage degree, bare rock ratio of these ten parameters space distribution ratio, and accumulate experience for the further analysis of the influence of Karst Rocky Desertification.