4D holographic video capturing method

Abstract

The invention provides a 4D holographic video capturing method. Image information of a scene and an object are captured through a multi-camera image sequence synchronous acquisition system. Through using an image processing method and a computer vision processing method, an accurate 3D object model which corresponds with each frame of a photographed object is established according to each frame ofimage data and a camera parameter. Rendering reconstruction is performed on the texture of the photographed object through texture rendering, thereby obtaining a vivid reproduced three-dimensional photographing object and a photographed object three-dimensional model of which each frame can be viewed in different angles. A 4D holographic video is obtained through continuously playing the frames,thereby reproducing a three-dimensional static or dynamic object in a high vivid manner.

Description

technical field [0001] The invention relates to the technical field of video imaging, in particular to a 4D holographic video capture method. Background technique [0002] 4D video includes four dimensions, three-dimensional spatial dimension and one-dimensional time dimension. 4D holographic capture refers to a high-quality and realistic 4D virtual reproduction method for static or dynamic scenes in real space. Existing 3D movies usually form a 3D stereoscopic viewing experience from the user's viewing angle by forming left and right eye imaging views respectively from the user's viewing angle. In fact, it only obtains 2.5-dimensional depth information without restoring it Get a true 3D view of your subject. [0003] High-quality real-time 4D video has many potential applications in virtual reality (VR), augmented reality (AR) or mixed reality (MR). For example, it can be used for character creation in 3D games and 3D animation. In the field of clothing, 4D video can rea...

AI Technical Summary

Problems solved by technology

[0004] Although various VR panoramic cameras, VR glasses, helmets and other equipment and products have appeared in the VR market at home and abroad, the VR content that can be viewed is very scarce, which limits the promotion of related products and applications.

The existing portable panoramic camera can shoot 360 videos, but the position and movement of the main angle of view in the 360 ​​videos are determined by the photographer and cannot be completely determined by the user

And the resulting 360 video isn't really a VR video

At present, 3D human body models can be created by using 3D software such as 3D Max, Maya and other tools, but the creation process of this kind of CG model is complicated, the fidelity is low, and the image distortion is serious, which cannot meet the requirements of real VR video

None of the current methods can meet the requirements of VR video and realistically create subjects in virtual or hybrid worlds in 3D

[0005] At present, the 3D reconstruction based on laser scanning, structured light or stereo vision can only be used for static scenes, and the accuracy and fidelity of the reconstruction are poor; the depth information of objects can be obtained by using depth sensor devices such as kinect, and the fusion of multiple kinect data The method can also obtain 3D information, but the method based on kinect is limited by the effective working distance of kinect, and it is impossible to shoot objects that require a large movement space

Moreover, the existing 4D holographic video capture methods cannot efficiently realize 4D capture for dynamic scenes.

Images