Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device

GPU-based video editing effects technology Research and Implementation

Author LuZhiYuan
Tutor CaiHongBin
School University of Electronic Science and Technology
Course Computer Software and Theory
Keywords Nonlinear Editing Graphics Processor Video Effects Binocular video
CLC TP391.41
Type Master's thesis
Year 2009
Downloads 136
Quotes 1
Download Dissertation

In recent years, computer graphics technology, the rapid development of digital video technology matures, the digital video processing core nonlinear editing video processing systems have become production of television programs, the main tool for digital multimedia. In traditional non-linear editing system, effects are usually based on the CPU or a dedicated board to achieve. However, when the image resolution is large, the image too much, the more complex effects, excessive consumption of the CPU or the board can not be achieved, resulting in the breakdown or video transition is not smooth, and it works with our enhanced visual effect contrary to the original intention. And the price is too expensive dedicated boards. In recent years, the rapid development of computer technology, there has been the central processor-based graphics processor input and output boards (CPU GPU I / O) technology, no card nonlinear editing system, and gradually become the market mainstream products. Video editing is non-linear editing effects technology is an important technology in technology, to enhance the visual effect of video editing, creative personnel to better express intent. Video editing effects technology, also known as digital video screen stunts, stunt short video or digital effects, is in television production, the use of digital video production equipment, video screens for special processing techniques, in order to achieve a certain method of making the picture effect . Any use of digital video production technology, in the video editing, production process, the video screen for special treatment of the various techniques, video editing effects technology belong to the category. In this thesis, the video editing effects technology, including for ordinary monocular and binocular video video processing, and ultimately used AVS nonlinear editing systems. Firstly, the graphics processor technology, video editing systems, system development platform, texture mapping and other video editing effects related technologies conducted in-depth research and analysis, followed by details of the non-linear editing system, a variety of video editing effects (including linear transformation , nonlinear transformation, fuzzy processing, scaling, mirroring, keying, a variety of wipe, fade, flip, page curl, etc.) of the algorithm, and gives the key functions to achieve stunt f (x, y), with switching the control law of each frame, in order to achieve different effects. Then introduced the system architecture, function interface, implementation, and based on the GPU to achieve these effects, while achieving binocular video editing effects. On an experimental basis, the paper also presents a video editing trick when generating data processing acceleration algorithms. The main innovation is that all video editing effects are based on the GPU to achieve, improve the efficiency of video editing, video editing and special effects to achieve binocular.

Related Dissertations
More Dissertations