Outdoor three-dimensional scene combined construction method based on image content parsing

Abstract

The invention provides an outdoor three-dimensional scene combined construction method based on image content parsing. The outdoor three-dimensional scene combined construction method comprises the following steps of: acquiring scene layout information by parsing an image map, acquiring geometric structure information of scene objects by parsing a plurality of internet images of a scene, and combining the scene layout information with the structure information of the objects to obtain a description of three-dimensional information of the integral scene; and constructing a three-dimensional object material library, retrieving a model in the library by semantic and model feature information, converting each object or object part in the described scene into a specific three-dimensional model, carrying out operations of deformation and the like according to the layout and structure information of the scene, and finally, obtaining a combined construction result of the integral outdoor three-dimensional scene . The three-dimensional scene description method provided by the invention can effectively describe the layout, structure and semantic information of the three-dimensional scene and provides basics for retrieval; and on the other hand, the invention also provides a model retrieving and matching reference method.

Description

Technical field [0001] The invention relates to the field of computer vision and three-dimensional scene modeling, in particular to an outdoor three-dimensional scene combined construction method based on image content analysis. Background technique [0002] Three-dimensional scene modeling is one of the important issues in the computer field. It has been widely used in virtual photography and film and television post-production and has played an important role. [0003] The use of virtual photography can be traced back to the Cold War. At that time, some countries developed battle simulation systems based on the needs of national defense training, which can be described as the predecessor of virtual studios. After the Cold War, because there was no need for war, this system was improved and launched. The advantage of virtual photography is that the space can be used effectively and can be reused. A virtual studio can build a scene larger than the actual space, and different scen...

AI Technical Summary

Problems solved by technology

However, traditional 3D scene modeling is generally done manually by art designers, which is time-consuming, labor-intensive and slow

Images