Research On Spatial Relation Reasoning Method Based On 3D-ICSRM Model

The existing composite models of three-dimensional space had some shortcomings in describing the position relationship either in cognitive habits or discrimination aspects between the target objects,these defects directly affect the accuracy of spatial reasoning results.Aiming at the above problems,this paper proposes a composite expression model,and a high precision spatial relation reasoning algorithm based on this model is given.The main contents are as follows:Firstly,in order to improve the accuracy of the spatial object position information description.Combining the advantages of the eight direction cone model and the 5 qualitative distance division method,a three-dimensional compound expression model(3D-ICSRM)is proposed.The position information between the 3D object is defined and described in detail based on this model,and the process of constructing 3D-ICSRM model is given.The performance of the model is evaluated according to the performance indexs of the space model.Secondly,the reasoning results of the position relation between the target objects is given by the compound tables method.In order to improve the accuracy of the reasoning results and narrow the results to find the scope,a high-precision spatial reasoning algorithm(HPSRA)is proposed According to the interaction mechanism,different spatial relationship of three-dimensional coordinate transformation principle and calculation formulas,which combines qualitative representation and quantitative derivation.The results and analysis of the combination reasoning between different spatial target objects are given by the above algorithm.Finally,the spatial reasoning prototype system is established.The image simulations of 3D-ICSRM model and spatial reasoning results are given,and the correctness of the proposed model and inference algorithm is verified.Compared with the existing models,the 3D-ICSRM model is proved to have good performance in terms of indexing and reasoning,As well as the advantages of HPSRA algorithm in terms of accuracy,error and so on when compared with compound tables method is proved.