Research On The Similarity Measurement Of Multi-scale Spatial Objects

With the rapid advancement of spatial data acquisition methods and technologies,a large number of basic geographic databases have been gradually built and put into use.There is a contradiction between existing spatial data and the rapidly changing real world.The instantaneity of spatial data has become increasingly prominent.Currently,the geoinformatics community generally agrees that multi-scale map data linkage updating technology is an important means to improve the spatial data's instantaneity.The key is the object matching that is establishment of the corresponding relationship between the same-name space objects on different scale maps.The same object on maps of different scales and different sources usually has certain differences.In order to evaluate the object matching results,it is necessary to measure the similarity of the spatial objects from different scales in the matching object pairs.The existing similarity measure methods mainly include the compactness method,tangent space function method,two-parameter method,Fourier descriptor method,shape multi-level description method,etc.However,these methods only consider the object global distribution or local structure.For this reason,this paper combines the characteristics of the description of the fractal dimension in fractal theory with the characteristics of rotation,translation and scaling,and proposes a similarity measure method of multi-scale map based on the fractal dimension.For the line object,Fractal dimension-based multi-scale line elements similarity measure method uses the fractal dimension to describe the geometric features of the line elements,and combines its distance to quantify the similarity between multi-scale line elements with the same name.For the polygon object,fractal dimension-based multi-scale polygon elements similarity measure method uses the fractal dimension to describe the geometric features of the polygon elements,and combines its spatial position,size and distribution mode to quantify the similarity between multi-scale polygon elements with the same name.This method not only takes into account the local structure of the object,but also takes into account the global distribution of the object.Finally,this method was applied to the similarity measure of point group object,and achieved better results.The research content of fractal theory is the system of irregular geometric shapes and widespread disorder and self-similarity in nature.Among them,self-similarity is an important concept in the theory.That is the overall reduction will be the same as the local one,and the partial enlargement will be the same as the whole.Fractal dimension is the most important concept and content in fractal theory.It can express intuitive morphological features,and can provide a good mathematical method for the study of irregular and complex elements.According to the characteristics of fractal dimension,it takes into account the local structure and global distribution of the object.It is consistent with human visual perception and is a better method for measuring the similarity of shapes.Therefore,the fractal dimension is introduced into the multi-scale spatial object similarity measure model,and the practicality of the similarity measure method based on fractal dimension is verified through experiments.Through the research on multi-scale spatial object similarity measurement model,this paper provides the basis for the spatial data quality evaluation and comparative analysis,promotes the improvement of the spatial data quality,and can also improve the efficiency of automatic database updating.