Method and device for carrying out 2D to 3D conversion on image

Abstract

Embodiments of the invention disclose a method and a device for carrying out 2D to 3D conversion on an image. The 2D to 3D conversion efficiency and effect of the image can be improved. The method comprises the steps of S1, acquiring a to-be-processed 2D image, inputting the to-be-processed 2D image into a pre-built and trained parallax information extraction model to obtain a parallax informationimage; and S2, carrying out three-dimensional rendering on the to-be-processed 2D image in combination with the parallax information image, and performing three-dimensional reconstruction on the to-be-processed 2D image.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of three-dimensional display, and in particular to a method and device for converting an image from 2D to 3D. Background technique [0002] The process of converting traditional artificial 2D content into 3D mainly includes the following steps: [0003] 1. First of all, it is necessary to perform roto processing on the objects in the image, and draw the edges of all objects in the screen with coils, so as to distinguish different object areas. This technology requires long-term training for technicians who can skillfully describe objects in the screen through multiple coils. The accuracy of object edges directly affects the quality of the final converted 3D content. [0004] 2. Obtain the parallax image corresponding to the 2D image. Although there is no depth information in the 2D image that can be recognized by physiological stereo vision such as binocular parallax, there are depth h...

AI Technical Summary

Problems solved by technology

[0007] To sum up, the production process of manual 2D to 3D content is complex, the personnel training period is long, and the cost of 3D content conversion is extremely high. The conversion process is completely manual. directly affects the quality of the final 3D reconstruction

[0008] Another low-intelligence 3D conversion technology, such as the one-click 3D conversion function of 2D video launched on YouTube, mainly uses color to distinguish the front and back, rather than imitating the real space structure, and the error rate is extremely high. However, the MIT and Qatar Computer Research Institute proposed A kind of real-time 3D conversion limited to the football screen by extracting data from video football games, but due to the large limitations and the stereoscopic effect segmentation is not obvious

Therefore, the problem of shortage of 3D content cannot be solved, which seriously hinders the development of 3D channels and 3D terminals

Images