Study On Fast Super-resolution Restoration Technology Of Video Image

Along with the development of digital media technology, a large number of video andimage information influx into people’s daily lives and various fields of the nationaleconomy，and the people would like to obtain more and more high-resolution images. Thesuper-resolution image restoration technology can improve the image’s resolution and getmore details and information without changing the existing imaging equipment. So itbecomes one of the hot research areas in recent years.The development of super-resolution enhancement technology depends on thesuper-resolution enhancement algorithm and its hardware implementation.They are distinct,self-contained but closely related.The practical application is the driving force to promotethe research and development of super-resolution enhancement algorithms. The purpose ofimplementing the super-resolution enhancement algorithms in hardware is to enhance theprocessing speed to reach the real-time effect and to prepare for a system on chip. As anintegration of embedded processors and FPGA, P-SOC provides the possibility of makingSuper-resolution video restoration system for its characteristic of collaborative design ofboth software and hardware as a FPGA chip which has integrated embedded processors.On the foundation of the existing research of super-resolution algorithm, this paperselects a super-resolution image restoration algorithm based on wavelet transformation andwrites c programs. The experimental results show that this algorithm has better effects ofrestoration than interpolation.Based on these works, this paper builds a super-resolution video restoration system onthe platform of zedboard development board from Digilent. This paper completes videocapturing of USB camera by using V4L2, and it designs graphical user interfaces by QT todisplay the video. Finally, a video capture-restore-display system on the development boardwas completed and the time spent on processing super-resolution algorithm on zedboardand Tiny210was compared. The experimental results show that this system is better thanTiny210platform in speed.