Adaptive Simplification Method of 3D Mesh Model Based on Local Area Features

Abstract

The invention discloses a method for self-adaptive simplification of a three-dimensional grid model based on local area features, comprising steps: 1) constructing a local area data set from a three-dimensional model data set; 2) training an MLP classification network: using the local area data set to train a Classification network based on local area features and simplification rate; 3) Calculate vertex normal deviation threshold: define the functional relationship between the classification result and vertex normal deviation threshold, and use the classification result to calculate the vertex normal deviation threshold. The present invention proposes a local area extraction method based on the topological structure of a three-dimensional grid model, using MLP to train a classifier for regional features, and when simplifying, classify each local area and use the classification result to guide the vertices of the local area to set the normal deviation threshold, That is, different simplification termination conditions are set for local areas, which avoids setting a uniform simplification rate for the model, realizes adaptive simplification, and obtains a higher simplification rate.

Description

technical field [0001] The invention relates to the technical field of simplification of a three-dimensional grid model, in particular to an adaptive simplification method of a three-dimensional grid model based on local area features. Background technique [0002] With the development of computer graphics and the improvement of computer performance, 3D models have been widely used and researched in fields such as virtual reality, animation games, and manufacturing. The rapid development of 3D data acquisition and modeling technology has made the accuracy of 3D models higher and higher, and the amount of data has also increased rapidly, which has brought huge pressure to the computer's drawing, transmission, editing and other systems. One way to solve these problems is to simplify the 3D model. At present, the commonly used grid simplification methods mostly control the simplification process from the geometric level. By setting a global simplification rate, calculating the...

AI Technical Summary

Problems solved by technology

However, the existing mesh simplification algorithm considers the geometry, topology information, color, texture and other attributes of the model to constrain the simplification process, and rarely considers the problem of adaptive simplification according to the characteristics of different regions of the model.

Images