Virtual and real scene mutual shielding realization method

Abstract

The invention discloses a virtual and real scene mutual shielding realization method. The method comprises a method for obtaining virtual and real scenes, a method for decomposing virtual and real scene images, a method for introducing an alpha channel, and a method for finally realizing virtual and real multilayer shielding. The virtual and real shielding method is realized under the condition ofno need of real scene prior information. Firstly, the virtual and real scenes are obtained; then the virtual and real scene images are decomposed by applying an existing motion target extraction technology in a complex background; and the Alpha channel characteristics of the images are fully utilized, so that multilayer shielding between virtual and real objects is realized. According to the method, the real scene does not need to be known in advance and modeled, so that the early-stage workload is greatly reduced; the method can be flexibly applied to any real environment and has relativelyhigh universality; and in addition, the motion target extraction technology is used in the method, so that the shielding phenomenon of non-rigid objects also can be well realized.

Description

technical field [0001] The invention relates to the field of augmented reality technology, in particular to a method for realizing mutual occlusion of virtual and real scenes. Background technique [0002] Augmented Reality (AR—Augmented Reality) is an important new means of realizing multimedia technology in the three-dimensional field. Augmented reality is a technology that uses three-dimensional information generated by computer systems to increase users' perception of the real world. In the augmented reality system, realistically rendering the mutual occlusion relationship between virtual and real objects is an important prerequisite for truly realizing the seamless fusion of virtual and real, and will also effectively improve the visual consistency of the system; [0003] In order to make users feel convinced that the virtual environment is a part of the real environment around them, the system should provide users with a correct sense of depth. The occlusion relations...

AI Technical Summary

Problems solved by technology

[0004] Although researchers have invested in the research of augmented reality technology since the 1990s, there are relatively few research reports on the virtual and real occlusion of AR systems that can be consulted. Over the years, researchers seem to have paid more attention to Focusing on issues such as tracking registration and 3D reconstruction, researchers at home and abroad have proposed different occlusion solutions for the video see-through and optical see-through AR systems, each of which has advantages and disadvantages;

[0005] At present, there are few researches on AR virtual and real occlusion in China, and although some methods have been proposed abroad, most of them are based on the idea of ​​3D modeling of all objects in the real scene, which requires that all information in the real environment must be known in advance. When the object moves, the pre-made model of the object is used to achieve occlusion. The problems of this type of method are: (1) 3D modeling of the real scene requires a lot of tedious preliminary work; (2) due to the use of advance Make a good model of the real object to achieve occlusion, so once the real scene model is made, any change in the real environment may cause the failure of the system occlusion implementation, so the portability of this AR system is poor; (3) For real scenes For non-rigid objects, such as people, the change in body shape when moving cannot be accurately predicted in advance, so the non-rigid object models made in advance also lack accuracy or even vary greatly, resulting in occlusion failure; optical see-through AR The system has the advantages of simplicity, high resolution, and no visual deviation, but it also has disadvantages such as high positioning accuracy requirements, difficult delay matching, relatively narrow field of view, and high price.

Images