Analysis processing method and device of three-dimensional model

Abstract

The present invention provides an analysis processing method and device of a three-dimensional model. The method includes the following steps that: three types of adjacent relationship information ofeach part in the three-dimensional model are determined; all parts of the three-dimensional model are aggregated, so that three component hypothesis sets of the three-dimensional model are generated;voxelization expression processing is performed on each component in each component hypothesis set, so that the voxelization information of each component hypothesis is generated; the voxelization information of all the component hypotheses in each component assumption set is inputted into a recognition model, so that the credibility scores and semantic category probability distribution of each component hypothesis are obtained; and a high-order conditional random field model is established according to the credibility scores and semantic category probability distribution of each component hypothesis in the three component hypothesis sets, and the semantic category analysis result of each part in the three-dimensional model is obtained. With the method and device adopted, analysis processing can be performed on the semantic categories of a three-dimensional model of an internal-external structure, and good analytical results can be obtained.

Description

technical field [0001] The invention relates to the technical field of virtual reality and augmented reality, in particular to a method and device for analyzing and processing a three-dimensional model. Background technique [0002] With the development of computer graphics, virtual reality technology and augmented reality technology, the application of 3D models is becoming more and more extensive. It is necessary to perform semantic analysis on the 3D model, that is, it is necessary to know which semantic category each part in the 3D model belongs to, wherein each semantic category is a structural category in the 3D model. For example, it is necessary to know whether the parts in the 3D model of the car belong to the semantic category of the door, or the semantic category of the front, and so on. [0003] In the prior art, a projection-based 3D model semantic segmentation method is adopted, which proposes a novel two-type Hausdorff distance to match the binary images proj...

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a method and device for analyzing and processing a three-dimensional model, which is used to solve the problem in the prior art that the analysis and processing of semantic categories cannot be performed on a three-dimensional model with an "inside-outside" structure, and the effect of analysis and recognition is poor

Images