Font Size: a A A

Study On Artificial Forest Dynamics Using Object-based Image Analysis At Shapotou District,Zhongwei City, Ningxia Autonomous Region

Posted on:2013-06-07Degree:MasterType:Thesis
Country:ChinaCandidate:X M ShaoFull Text:PDF
GTID:2230330371986584Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
The article extracted the artificial forest using object-based image analysis at Shapotou District, Zhongwei City, and achieved the artificial forest dynamics change detection in the past20years based on overlay analysis of GIS. Finally, the article analyzed the driving force of the reasons for the changes. The conclusions are as follows:(1) Compared with the traditional pixel-based image analysis methods, the object-based image analysis was more in line with Person’s logical thinking habits, and improved the efficiency of processing and analysis. In addition, it increased the spatial analysis capability effectively. The object-based image analysis had great advantages.(2) This article evaluated the image segmentation results using the inconsistencies evaluation method. We proposed three new metric indexex on the basis of integrating geometric consistency and algebraic consistency. They were PSE、 NSR and ED2. We calculated the indexes, and selected the best segmentation parameter combination by the calculation results. In order to reduce the change detection error caused by the segmentation error, the article united the best segmentation parameter combination of each phase of remote sensing images.(3) This article selected many fetures, such as DVI, RI, Mean B5, Mean B6, Length/Width, GLCM Mean B2, Distance to line, DEM etc, as the parameters for the rule set. Then we selected the threshold method or member function to complete the establishment of the rule set according to the distribution of surface features’eigenvalues.(4) From2000to2010, the spatial distribution of timber forest was expanding constantly. During the entire study period, the rate of timber forest’s change was from ’sharp’to’slow’. From1992to2003, the area of protection forest had growing trend. But from2003to2010, its area had a certain degree of reduction. The rate of protection forest’change fluctuated between ’slow’ and ’very slow’ during the entire study period.(5) There were five kinds of land change type in the study area, such as constant other surface features, constant protection forest, into timber forest, into protection forest, and fluctuating. The covering area of’constant protection forest’ distributed mainly on both sides of the Baotou-Lanzhou Railway. Its quality of the ecological environment was relatively stable. The covering area of’into timber forest’and’into protection forest’distributed in the southern margin of Tengger Desert and the local area of the farmland oasis at Shapotou District. To a certain extent, its quality of the ecological environment was improved. The covering area of’fluctuating’distributed on both sides of the railway and in the region of timber forest developing. Its quality of the ecological environment was fluctuating.(6) In the four study periods, the timber forest had inflows all. And the main inflow types were the flow sand and the steppe desert. In the two study periods of2003-2006and2006-2010, the timber forest had outflows. The main outflow types were the urban and rural construction land. The protection forest had inflows in the two study periods of1992-2000and2000-2003. The main inflow types were the steppe desert and the flow sand. In the two study periods of2003-2006and2006-2010, the protection forest had outflows. The main outflow type was the steppe desert.(7) In the past20years, the sand-fixing activity of human was the decisive factor and the main driving force for the artificial forest changes in the study area.
Keywords/Search Tags:object-based image analysis, image segmentation, rule set, dynamicdegree, state transition matrix, driving force, artificial forest, Shapotou
PDF Full Text Request
Related items