随机几何应用于遥感数据地物目标提取方法综述

Research on Application of Stochastic Geometry to Ground Objects Extraction from Remote Sensing Data

遥感数据地物目标(如道路、建筑物、水系和树冠等)提取有利于交通管理、城市规划等相关领域的发展。随机几何理论应用于目标提取具有灵活性、面向对象性、稳定性和收敛性等优势,逐渐成为学者研究遥感数据地物目标提取方法依据的重要理论基础。鉴于此,依据近些年来国内外的相关文献,对现有的随机几何理论和方法进行了总结,梳理了随机几何理论应用于地物目标提取建模的基本思想,介绍了该类模型的参数求解算法以及实验结果的评价方法,并针对现有方法及其特点总结了随机几何理论应用于地物目标提取的建模流程。总结了现有方法存在的优势与劣势,并从提取效率、精准性、应用性和求解算法完备性四方面给出了未来研究的可能性,为随机几何理论的应用以及遥感影像解译提供了有价值的思路。

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.