Research on Application of Stochastic Geometry to Ground Objects Extraction from Remote Sensing Data
Ground objects(e.g.,roads,buildings,water systems,tree crowns,etc.)extraction from remote sensing data is beneficial for the development of related fields such as traffic management and urban planning.The application of stochastic geometry theory in target extraction has advantages of flexibility,object-oriented,stability,and convergence,thus it is gradually becoming an important theoretical basis for scholars to study the methods of extracting ground objects from remote sensing data.Accordingly,based on domestic and foreign references in recent years,the existing theories and methods are summarized,the basic idea of applying stochastic geometry theory to the modeling of ground object extraction is analyzed,the parameter solving algorithm of this kind of model and the evaluation method of experimental results are introduced,and the modeling process of ground objects extraction is drawn.Finally,according to the advantages and disadvantages of existing methods,the possibilities of future researches from four aspects,i.e.,extraction efficiency,accuracy,applicability and completeness of solving algorithms are given and valuable ideas to continuously realize theoretical and methodological innovation in related fields are expected to provide.
stochastic geometryremote sensing datapoint processsimulation algorithmground objectsline process
国家科技期刊平台
NSTL
万方数据
