Video splicing method based on Bayesian theory

A technology of Bayesian theory and video splicing, which is applied in TV, color TV, color TV parts, etc. It can solve the problem of unresolved character movement details and achieve the effect of ensuring stability

Active Publication Date: 2016-06-01
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF2 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Traditional optical flow vector estimation requires dense sampling. Of course, the transformation from coarse to fine reduces this limitation to some extent, but the details of character movement have not been resolved.

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
  • Video splicing method based on Bayesian theory
  • Video splicing method based on Bayesian theory
  • Video splicing method based on Bayesian theory

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] Describe technical scheme of the present invention in further detail below in conjunction with accompanying drawing: as figure 1 As shown, a video stitching method based on Bayesian theory, which includes the following steps:

[0031] S0: The video frame is preprocessed, and the image is corrected so that the center of the image is on the same horizontal line.

[0032] In the process of panoramic image stitching, image correction is a very common step, through image correction to make the center of the image on the same horizontal line. We usually compute their positions based on camera calibration. The rectified image planes can accurately appear in pairs in the input image. If the camera parameters are unknown, various techniques for extracting the geometric epipolar lines can be applied.

[0033] The geometric relationship between cameras, three-dimensional space and images, C and C' represent a pair of cameras in three-dimensional space; w and w' are the correspo...

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 provides a video splicing method based on the Bayesian theory. The method comprises steps of estimating a flow field for a corresponding frame of a video, and considering image registration errors caused by inaccuracy of the flow field; applying the Bayesian theory in time and space and using the image errors as self-adaption reliability so as to obtain a posterior probability model after splicing of the video; and solving a maximum posterior model by means of the optimization method so as to obtain stable high-quality video output results. According to the invention, singular points caused by inaccuracy of estimation of the flow field can be effectively removed in space and stability of the video is ensured in time, so high-quality video output results can be obtained.

Description

technical field [0001] The invention relates to a video mosaic method based on Bayesian theory. Background technique [0002] Video splicing is to give a series of video image sequences with limited overlapping areas, process them through algorithms, and stitch them into a high-resolution, wide-angle panoramic video image. Generally speaking, panoramic video stitching technology can be divided into two categories according to different acquisition methods: one is obtained through a single special camera lens Panoramic video images; the other is to obtain videos from different perspectives through ordinary cameras, and then obtain panoramic images through stitching algorithms. Although the former can quickly obtain panoramic video images, it is expensive and the output video quality is not high; the latter has fewer constraints on the scene and the photographer, and is more practical, so more users use video streams from multiple perspectives Create panoramic video images....

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 Applications(China)
IPC IPC(8): H04N21/44H04N5/232
CPCH04N21/44016H04N23/68H04N23/698
Inventor 杨路苏建美王泽程洪
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap