3D model feature extraction method based on compressed sensing

Abstract

The invention provides a 3D model feature extraction method based on compressed sensing. The method comprises the following steps: first, selecting a 3D model in a discrete voxel format, selecting the orientation of each angle as a reference plane, designing an equal-altitude transformation function, and spatially layering the 3D model according to the equal-altitude transformation function; then, projecting each spatial layering model onto the reference plane, constructing projection matrixes, and extracting the information entropy of the projection matrixes; and finally, sparsely processing each projection matrix, and carrying out 2D compressed sensing treatment to get spatial layering features. Through the method, the features of a 3D model can be reflected from multiple angles, spatial layering of a 3D model in a voxel format is realized, and a complex 3D model can be spatially decomposed. The accuracy and efficiency of 3D model feature extraction are improved. Moreover, low-dimensional efficient space geometric features are extracted, and feature redundancy is avoided. Thus, the speed and quality of 3D model retrieval are ensured.

Description

technical field [0001] The present invention relates to the field of three-dimensional model processing, and more specifically, relates to a three-dimensional model feature extraction method based on compressed sensing. Background technique [0002] With the rapid development of information retrieval technology and the improvement of computer performance, information processing has changed from traditional mode to new mode. Compared with text information and two-dimensional images, more realistic and rich three-dimensional models are more and more widely used. In today's massive 3D model database, how to realize the management and retrieval based on 3D model reuse, and quickly and accurately find the 3D model that meets the requirements has become an important research topic in the retrieval field. [0003] As the fourth multimedia data type after sound, image and video, the development of content-based 3D model retrieval technology has attracted much attention. How to qui...

AI Technical Summary

Problems solved by technology

[0004] At present, most of the content-based 3D model retrieval methods still have some problems: such as the extracted features cannot fully express the 3D model information, the computational complexity is high, the time for feature extraction and feature matching is long, the feature storage space is large, feature information is easily missing, and cannot Realize user interaction, etc.

Images