Implementation of Basic Image Processing Applications by FPGA Hardware

Abstract

At present, the relevancy to image processing is increasing and applications of image processing are developing. Also ‘Field Programmable Gate Arrays (FPGA)’ is gaining in popularity nowadays. FPGAs are strong in parallel computation and they can work too fast. Because of these advantages, they are preferred in applications which have complex calculations and need high speed. At the first part of this work, an image processing application was implemented on FPGA hardware by using ‘diamond-square’ algorithm and several image processing techniques. For this application, it isn’t necessary to use camera. Because image is generated by using algorithm, then image processing techniques like filtration and rotation are implemented to image. At the second part of work, real-time video motion was used for basic image processing applications. Image was taken by using 5 Megapixel camera, then gray-scale and binary images were obtained. Also Sobel edge detection algorithm was implemented to image with variable threshold.  All of these algorithms are obtained by using Verilog hardware description language. Results of applications and information about the performance of hardware were given at the result part of work.

Keywords: diamonds-square algorithm, edge detection, gray-scale image, binary image, Verilog, FPGA.