| Karst rocky desertification is a kind of special and complicated surface form,which is performance for the fragile ecological environment,and lead to the deterioration of regional ecological environment,so that drop the effective use of land resources, water resources have been destroyed, the survival of the foundation of the loss of biological resources.Remote sensing technology is an important means for timely and accurate access to geographic information,with a wide range of detection, the amount of information rich, comprehensive, periodic dynamic monitoring and other characteristics,which have became the main method of investigation and monitoring of rocky desertification.The research of traditional optical remote sensing technology is basically mature in being applied to the investigation and monitoring of rocky desertification,but in the Guizhou plateau as the center of China’s southern Karst region due to cloudy and rainy climate,optical data is difficult to be used as the monitoring data of Karst rocky desertification,and there is a large range of data blind area.Microwave remote sensing has the limitation of weather and illumination conditions,compared with optical remote sensing, the characteristics of surface illumination can be obtained,polarization SAR has the advantages of electromagnetic scattering vector and microwave penetration,full polarization SAR will become Karst Rocky desertification Remote Sensing Investigation and monitoring of another data.Study selection Radarsat-2 full polarimetric SAR data as remote sensing monitoring data,in Liu Chang Qingzhen City of Guizhou province base unit based on modern tobacco agriculture,the establishment of Karst rocky desertification grade classification index system was established from the terrain factor, lithologic factor, vegetation factor and human factor,use 1:1 million topographic map to establish DEM as the terrain data,the lithology distribution is taken as a lithologic factor by the vector distribution of 1:20 ten thousand,the vegetation cover is obtained by the method of polarization decomposition and mixed pixel dichotomy,Based on the target decomposition, the H/ α space of the target object is extracted,analysis of the scattering mechanism of land use in the study area as a human factor;the decision tree and the comprehensive analysis method were used to monitor the Rocky Desertification in the study area,the information of Karst rocky desertification classification was extracted in the study area,and mapping the Karst rocky desertification grade classification.Results show that investigation on Karst rocky desertification remote sense image data and Radarsat-2 monitoring has feasibility and accuracy,which can get real time data of Karst area and it can provide a new technical scheme for the investigation and monitoring of ground objects identification in Karst rocky desertification area,to provide data support for the ecological security and ecological restoration in rocky desertification area.The main conclusions of the study are as follows.(1)SAR data preprocessing.The research use the Radarsat-2 full polarimetric data as a remote sensing investigation and monitoring data source,a SAR data processing procedure is established for Karst mountainous area.Using the ENVI, PloSARpro, Erdas and other remote sensing data processing platform on study, formed a set of raw data pretreatment process for the Karst mountainous area in South China,Which include the multi view processing,coherent matrix extraction,wave filtering,feature parameterextraction and geometric correction.(2)Extraction of rocky desertification index based on SAR polarization decomposition.On the basis of extracting characteristic parameters(RVI), the information of mixed pixel is extracted by using pixel two model,combining with vegetation coverage in the study area,according to the measured data and SAR image data,determined the RVImax is 1.024,RVImin is 0.142,extracted and calculated vegetation coverage in the study area;Based on a two-dimensional 2D space on H-α,using Wishart-H-α to extract land type,and carries on the precision verification.Research shows that the study area land use classification using Wishart-H-α supervised classification,its overall accuracy reached 83.13%,Kappa coefficient reached 0.81.Accuracy show that the Wishart-H-α classification method to get the land use types can be as Rocky of classification index system in human factors of rocky desertification interpretation.(3)Classification model and classification of rocky desertification degree.Based on the geographic information database of the established research area,using decision tree and comprehensive analysis method to establish the quantitative evaluation model of rocky desertification,obtained the rocky desertification distribution map,and the accuracy of the classification results is verified by the actual rock desertification distribution map.Studies have shown that the rocky desertification classification using decision tree method and c o m p r e h e n s i v e a n a l y s i s m e t h o d c a n o b t a i n b e t t e r c l a s s i f i c a t i o n accuracy,classification accuracy is respectively as 80.125% and 84.247%,Kappa coefficient is respectively as 0.7412 and 0.8141,but analysis accuracy is better than decision tree accuracy.Therefore, based on the full polarimetric SAR data, the degree of rocky desertification was divided, Research proposal select the comprehensive analysis method. |
PDF Full Text Request