Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

High-resolution monocular 2D video to binocular 3D video conversion method

A high-resolution, video technology, applied in image analysis, graphics and image conversion, instruments, etc., to achieve the effect of fast repair speed, clear and reasonable texture, and stable effect

Active Publication Date: 2022-07-12
SOUTHEAST UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0010] The present invention solves the above-mentioned problems and difficulties, and provides a method for converting high-resolution monocular 2D video to binocular 3D video. This method can process any high-resolution 2D video without camera parameters and other Under the premise of information, without manual operation, it can automatically convert the input monocular 2D video to 3D video, and maintain the original resolution and quality effect

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • High-resolution monocular 2D video to binocular 3D video conversion method
  • High-resolution monocular 2D video to binocular 3D video conversion method
  • High-resolution monocular 2D video to binocular 3D video conversion method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0083] Step 1: Extract the left and right frames from the left and right video of the 3D movie, and scale the resolution to 512*288. Then, calculate the texture color complexity C and brightness index L of the image, for and Y in image pairs with rich texture and color information and appropriate brightness are preserved and organized. and are the average texture complexity and brightness, respectively, which depend on the mean of the texture complexity and brightness of all extracted image pairs. h and w are the image dimensions, is the color component of the pixel in the (i, j) coordinate R channel, and the brightness Y is the brightness channel component in the YUV color space. The calculation rules are as follows:

[0084]

[0085]

[0086] Y=0.299R+0.587G+0.114B

[0087] The SGBM algorithm is used to perform binocular stereo matching on the image pair. As a global stereo matching algorithm, the SGBM algorithm can calculate the disparity map from the left a...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a method for converting a high-resolution monocular 2D video to a binocular 3D video. In the depth information extraction stage, a U-shaped convolutional network is trained by using an existing 3D movie as a source data set to obtain superior performance. The network model performs frame-by-frame depth estimation for 2D video, and uses a small neural network to optimize the edge-preserving smoothing of the depth map. In the viewpoint synthesis stage, a depth map-based viewpoint synthesis algorithm without camera parameters is proposed, and the left and right virtual viewpoints are synthesized by a symmetrical rendering strategy from the middle to the two sides. Finally, in the image inpainting stage, a block matching-based image inpainting algorithm combined with temporal information is proposed to fill in and repair the cracks and holes in the left and right viewpoints. The present invention can transform the original 2D video from 2D to 3D on the premise that the original 2D video does not have any relevant parameter information.

Description

technical field [0001] The invention relates to a method for converting high-resolution monocular 2D video to binocular 3D video, belonging to the fields of image processing, computer vision and computer graphics. Background technique [0002] In recent years, multimedia-related technologies and hardware devices have developed rapidly. Various 3D and VR devices have been launched on the market, such as VR glasses, 3D screen mobile phones and 3D monitors. At the same time, users' entertainment demands for watching 3D videos are also increasing. However, for a common mobile device with a shooting function, only common monocular 2D video can be shot, and there is currently a lack of technology for converting 2D video into 3D video on the market. Whether it is a 3D movie screen, VR glasses, or a 3D screen mobile phone and monitor, it needs to be supported by a variety of binocular 3D video resources to provide users with a 3D video viewing experience with stereoscopic, immersive...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): H04N13/261H04N13/268G06N3/04G06K9/62G06T3/40G06T7/593G06V10/774G06V10/82
CPCH04N13/261H04N13/268G06T3/4076G06T7/593G06N3/045G06F18/214Y02T10/40
Inventor 姚莉李林鑫杨俊宴吴含前
Owner SOUTHEAST UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com