Segmentation and semantic annotation method of geometry grid scene model

Abstract

The invention relates to the technical field of computer graphics, in particular to a segmentation and semantic annotation method of a geometry grid scene model. The method includes the following steps of building a three-dimensional training set, wherein each three-dimensional model in the training set is required to be a single object; automatically segmenting the scene model, wherein the scene model is segmented into multiple objects according to the training set and on the basis of the clustering hierarchy algorithm; classifying segmentation results, extracting shape characteristics of each object obtained through segmentation, and deciding a class label of the object according to the classification algorithm; collecting the semanteme of the scene model, and collecting the class labels of the objects to obtain a semantic label set of the scene model. Compared with the prior art, the method has the advantages that known shape knowledge in the training set is used in the automatic segmentation method of the scene model for assisting decision making. Therefore, the problem that contact objects are difficult to process during scene segmentation is solved, and semantic annotation of the scene model better fits visual perception of people for scenes.

Description

technical field [0001] The invention relates to the technical field of computer graphics, in particular to a segmentation and semantic labeling method of a geometric grid scene model. Background technique [0002] The 3D scanning technology is becoming more and more mature. The point cloud data of the 3D scene can be obtained by scanning equipment, and then the geometric grid representation can be obtained through 3D reconstruction for graphics rendering. However, the obtained 3D scene model lacks semantic information, and it is difficult for the computer to Understanding the content in the scene model makes it difficult to achieve content-based retrieval and effective management of the 3D scene model. The segmentation and semantic annotation of the 3D scene model is based on scene segmentation, that is, the scene is divided into multiple individual objects, and then the question of what is in the scene is answered, which has positive significance for the content understandi...

AI Technical Summary

Problems solved by technology

[0002] The 3D scanning technology is becoming more and more mature. The point cloud data of the 3D scene can be obtained by scanning equipment, and then the geometric grid representation can be obtained through 3D reconstruction for graphics rendering. However, the obtained 3D scene model lacks semantic information, and it is difficult for the computer to Understanding the content in the scene model makes it difficult to achieve content-based retrieval and effective management of the 3D scene model

[0004] From the above analysis, it can be seen that the segmentation of geometric mesh scene models has the following deficiencies: 1) most of the research on geometric mesh model segmentation is aimed at a single object, which is divided into several meaningful sub-components, while the segmentation of scene model There are few studies; 2) The existing methods for detecting and segmenting specific types of objects from the scene cannot be extended to the segmentation of various objects in the scene; 3) The method of using the scene graph to determine the objects in the scene is not universal, and many scene models Does not have a scene graph, or the scene graph is inaccurate

Images