Fpga Tabanlı Video Görüntü Çerçevelerini İyileştirme Cihazı Geliştirilmesi

Abstract

This dissertation aims to design hardware and software that ensure to restore and transmit the video frames in real time from the camera to the screen. Basically, a FPGA processing unit with the ability to perform parallel processing was used in the hardware, and the superresolution method in the software, in order to achieve this structure. In the simulation studies, an algorithm was created for motion estimation and reconstruction parts, which are the sub-units of the superresolution method. Integral projection model approach, which is the application of block matching algorithm, was used for the part of motion estimation. For the part of reconstruction, fusion of interpolation was used. The specified algorithm was created again using VHDL programming language so as to perform real time processing within the FPGA. Afterwards, the output frames were observed, transmitting the same picture frames to the input of each software, for testing purposes.