Research On 3D Mesh Model Segmentation Based On Improved Extreme Learning Machine

The 3D mesh model segmentation is to decompose the mesh into a certain number and connected regions according to the geometric or topological characteristics of the model,which makes the 3D mesh model segmented into several parts with simple shape significance.Therefore,3D mesh model segmentation is essential to the semantic understanding of 3D objects and is an indispensable basic research topic.The 3D mesh model segmentation method based on extreme learning machine,because the extreme learning machine randomly generates the input weight matrix and the hidden layer bias,the segmentation classifier model obtained by training is not good.In order to improve its shortcomings,this paper proposes two 3D mesh model segmentation methods based on improved extreme learning machines.The work done in this paper can be summarized as follows:(1)This paper proposes an evolutionary extreme learning machine optimized by ant lion optimization.This algorithm uses the ant lion optimization algorithm to optimize and improve the extreme learning machine.With the help of the special iterative update mechanism of the ant lion optimization algorithm,the elite ant lion representing the optimal solution is finally assigned to the input weight matrix and hidden layer bias of the extreme learning machine.Perform classification experiments on 5 types of UCI data sets to verify the feasibility and effectiveness of the proposed algorithm;(2)This paper proposes a 3D mesh model segmentation method based on optimizing extreme learning machine with ant lion optimization.ALO-ELM trains the feature data of the3 D mesh model to obtain a better segmentation classifier than the original,unoptimized extreme learning machine.And test the unsegmented model,and finally get the segmentation result with higher segmentation accuracy;(3)This paper proposes a 3D mesh model segmentation method based on optimizing extreme learning machine with fusion of differential mutation improved ant lion optimization.Use differential mutation to improve the ant lion optimization algorithm to avoid the situation where the elite ant lion will fall into the local optimal solution.Then,the improved ant lion optimization algorithm is used to optimize the extreme learning machine.And train the input3 D mesh model feature data to obtain a better segmentation classifier,which further improvesthe segmentation accuracy of the test model.