The Information Extraction And Evaluation Of Rocky Desertification In Chongqing City Based On The Association Of Unmanned Aerial Vehicle Photo And Landsat8 Image

Rocky desertification is a kind of desertification landscape phenomenon, which is caused by the common action of man-made destruction and natural conditions in the karst area of Karst, which leads to the destruction of vegetation, land degradation and rock exposed. It generally occurs in the southwest China carbonate rock area,it has become one of China’s southwest serious ecological environment problems, in view of the seriousness of the karst rocky desertification, in recent twenty years, the state has invested a large funds carry out rocky desertification evolution mechanism, prevention and control aspects of the research, the rocky desertification has been the initial containment. In order to assess the situation of rocky desertification after large-scale investment, the national or local organizations have carried out a number of large-scale rocky desertification investigation and evaluation, which provides a basis for the local government to control the rocky desertification. However, the traditional method of remote sensing is often used in the study of regional scale rocky desertification, and the data of rocky desertification is obtained by extracting the information related to rocky desertification. The precision of this method is closely related to the factors such as remote sensing image resolution, image quality, extraction method, and so on. But there is a lot of uncertainty. How to accurately retrieval the rocky desertification information by different kinds of remote sensing images with different spatial resolution is one of the difficult points to solve this kind of uncertainty.This paper take Chongqing city as the research object, through landsat8 remote sensing image, take high precision photo of aerial drones, sample surveys and other technical means, get different precision of rocky desertification information, combined with support vector machine model is of high precision machine data inversion and low precision landsat8 remote sensing the letter interest method, Chongqing rocky desertification status data, through analysis and field survey data contrast, the accuracy of this method can reach 97.4%, with images of landsat8 extraction of desertification data are compared, the accuracy is increased 24.1%. It provides a new method for the rapid monitoring of rocky desertification in the future. Through the research, the main conclusions are as follows:1. In ArcGIS using nets, alignment landsat8 TM remote sensing image pixel(30M * 30m), extracted pixel multi spectral TM image DN value and rate of UAV aerial image of the bare rock, using support vector machine(SVM) model of TM image multiple DN value and human image bare rock ratio in the regression analysis, and the model of remote sensing image of the bare rock rate, comparing with field survey data, inversion of the bare rock rate than simple remote sensing interpretation of bare rock ratio more close to the true value.2. Using the interpretation of the rocky desertification information, extract the relevant thematic maps, combined with the inversion of the bare rock ratio and field survey data, composite map, intuitive expression of the distribution of rock desertification.3. Analysis and evaluation of the present situation of rocky desertification, in Chongqing karst area 32734.42 km2, non rocky desertification area 14583.68 km2, accounting for 44.55% of the karst area. 10022.99 km2 of potential rocky desertification area, for the karst area of 30.62%; rocky desertification area 8127.75 km2, accounted for 24.83% of the karst area. Rocky desertification including mild rocky desertification 5851.43km2, moderate rocky desertification 2077.07km2, severe rocky desertification 199.25 km2, respectively, accounted for 71.99%, 25.56% and 2.45% of the occurrence area of rocky desertification.4. Analysis of the different grades of rocky desertification in different land use, the proportion of rock exposed rate, indicating the impact of these factors in the rocky desertification.5. Contrast rocky desertification monitoring data in 2005, Chongqing rocky desertification land area from 9256.6km2, reduced to now 8127.75km2. On the other hand, the potential rocky desertification land area increased from 8580.33km2 to 10022.99km2, the potential rocky desertification threatened the situation to become more severe.