CUDA (Compute Unified Device Architecture)-based panoramic video monitoring system

A technology of panoramic video and monitoring system, applied in the direction of closed-circuit television system, etc., can solve the problems of ghosting, difficult processing speed, and difficulty in ensuring real-time splicing, etc., to achieve the effect of improving efficiency and smooth real-time monitoring video

Active Publication Date: 2016-01-13
BEIJING UNIV OF TECH
View PDF11 Cites 41 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, for real-time video stream splicing, the processing speed of this algorithm cannot meet the real-time requirements, and also when the registration is not good, there will be ghost images
[0011] Due to the large number of pixels in the video image itself, the processing speed has always been a difficulty in video stitching. It is difficult to ensure the real-time stitching in the optimization of the algorithm stage.

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
  • CUDA (Compute Unified Device Architecture)-based panoramic video monitoring system
  • CUDA (Compute Unified Device Architecture)-based panoramic video monitoring system
  • CUDA (Compute Unified Device Architecture)-based panoramic video monitoring system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] Firstly, the phase correlation method is used to approximate the overlapping area of ​​the two images to be stitched. If image I 1 (x,y) and image I 2 There is a displacement relationship of (Δx, Δy) between (x, y), then the relationship between the two images can be expressed as:

[0032] I 1 (x,y)=I 2 (x-Δx,y-Δy)

[0033] The normalized cross power spectrum is defined as:

[0034]

[0035] in and Yes 1 (x,y) and I 2 Fourier transform of (x,y); complex conjugate. Inverse Fourier transform the above formula to get a two-dimensional function:

[0036] δ(x-Δx,y-Δy)=F -1 [e -j2π(uΔx+vΔy) ]

[0037] The position of the largest peak of the function corresponds to the translation parameter (Δx,Δy) between the two images. If there is only a translation relationship between the two images, the maximum peak value of the function reflects the correlation between the two images, and the value range is [0,1]. A value of 1 means that the two images are identica...

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 CUDA (Compute Unified Device Architecture)-based panoramic video monitoring system. Firstly, overlapping regions among a plurality of video images are estimated with a phase correlation method, and SURF (Speeded-Up Robust Transform) feature points are extracted in the overlapping regions and registered, so that algorithm time is shortened, and subsequent mismatching is reduced. Secondly, a fusion algorithm combining an improved optimal fusion line and a multi-resolution algorithm is provided, so that edge jumping and ghost phenomena are eliminated, and the visual effects of videos are improved. Lastly, a GPU (Graphics Processing Unit) is used for accelerating at a fusion stage, so that the splicing speed is further increased. As proved by experimental results, real-time splicing of three paths of monitoring videos can be realized effectively, and a frame rate is up to 20 frames through the system. Compared with a conventional CPU (Central Processing Unit) version, the system has the advantage that the real-time requirement of video splicing can be met.

Description

technical field [0001] The invention belongs to the field of video image processing and parallel computing, and relates to a panoramic video monitoring system based on CUDA. Background technique [0002] Human beings acquire a large amount of information every day, among which visual information is very important to human beings. Image is therefore very important as an important way for human beings to obtain visual information. However, as people have more and more requirements for scene information with a wider visual range, the angle and field of view of the scene that can be captured by ordinary cameras or video cameras are very limited, and in many cases it is necessary to obtain larger and more complete video images. getting information. Therefore, in order to obtain a wider viewing angle, multi-camera video stitching technology is extremely necessary. [0003] Video stitching is to simultaneously obtain video images with different positions and angles and partially...

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
IPC IPC(8): H04N7/18
Inventor 陶荷梦禹晶肖创柏段娟
Owner BEIJING UNIV OF TECH
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