Design And Implementation Of Super Resolution Image System Based On FPGA

As people’s requirements to the improvement of image resolution, the existing imaging device resolution are often unable to meet their requirements. In the digital image acquisition process, because of some artificial operation, people also can’t obtain satisfactory imaging. In order to obtain a higher resolution image in terms of imaging device improvements it need high cost, improving the image resolution method through the software attractes wide attention. Image Superresolution technology bases on analytic continuation and superposition of information management theory, attempting to restore the diffraction limit information and making the image get more details and information to improve resolution. The system bases on the FPGA (Field Programmable Gate Array, field programmable gate array), and it makes full use of hardware characteristics which can meet the need of high speed image processing. Coupled with the use of SOPC (System On a Programmable Chip) technology it allows the design flexible, meanwhile it can make hardware and software program and update on line.This paper describes the design of a FPGA based superresolution image system, using Altera’s DE2development board as hardware platform, SDRAM (Synchronous Dynamic RAM, synchronous dynamic random access memory) for image storage, VGA (Video Graphic Array, VGA) device displays images after Superresolution reconstruction.The whole system adopts software and hardware design method. Hardware:firstly, the system designs and realizes every module on the basis of overall demand analysis, including image acquisition module, SDRAM module, image display module and Superresolution algorithm module. By using QuartusII software and hardware design language this paper sets up the hardware system.When using QuartusⅡ, the system can use its SOPC Builder to build the Superresolution module as a follow-up analysis of operational platform. Algorithm:this paper bases on the image sequences Superresolution reconstruction algorithm. By the research of the algorithm principle and theory, this paper realizes the bilinear interpolation POCS(Projection Onto Convex Sets), MAP(Maximum A Posterior) in MATLAB. By analysis and comparision of the algorithm, this paper chooses high resolution image reconstruction algorithm which is based on MAP. Firstly this system deals with the image by gray-scale transformation and degraded transformation, and then progresses the low resolution image using interpolation algorithm and a priori knowledge and conditional probability to construct the objective function, lastly the image iterates with MAP estimation algorithm and gets a single image with more image information, more details. Software:in Nios II IDE environment the system uses C language to program and debugs the algorithm. After the debugging and running modifications it have been satisfied with the high resolution image.