3D Model Retrieval Based On Geometric Features

With the development of computer software and hardware technology,3D models are more widely used in fields of animation,machinery,medical and others,which results in increasing of the number of 3D models.Scholars are facing more challenges in the field of 3D model retrieval.Many scholars have researched on 3D model retrieval technology and proposed many valuable model retrieval algorithms,but there are still many problems that need to be solved urgently.This paper analyzed the existing technology of 3D model retrieval,and researched on topological structure features and shape distribution features of 3D model.In this paper,we proposed a 3D model retrieval method based on geometric shape features.For enhancing robustness of 3D model retrieval algorithm,geometric shape features with rotation invariance are applied to classification and retrieval of 3D models.Research contents in this article are shown as follows:In order to improve the efficiency of 3D model retrieval,this paper proposed a 3D model classification algorithm based on shape distribution feature fusion.For obtaining the global information of 3D model,D1 and D2 shape distribution features of 3D model are extracted.For obtaining partial information of 3D model,random spherical shape distribution features of 3D model are extracted.Auto-encoder is used to deal with these shape distribution features to obtain efficient features of 3D model.Finally,based on efficient features of 3D model,multilayer perceptron classifier is trained and is used to classify 3D models.Model database is used to verify the algorithm.Experimental results show that the method can effectively classify 3D models.This paper proposed a model similarity calculation method that fuses shape and structure information.According to the difference in the number of edges of3 D model faces,shape similarity between two 3D model faces is calculated.Shape similarity matrix of two 3D models is established based on it.According toadjacent conditions of faces,structural similarity between 3D model faces is calculated.Structural similarity matrix of two 3D models is established based on it.Based on shape similarity matrix and structure similarity matrix,shape-structure similarity matrix between two 3D models is established.Genetic algorithm algorithm is used to search shape-structure similarity matrix for obtaining an optimal face matching set.Similarity of two 3D models is calculated.Experiments show that the method can distinguish the difference of 3D models' topological structures.