Implementation of filtering algorithm on hardware platform

Abstract

This report covers the various aspect of hardware implementation on image filtering algorithm as well as the image structures which I have created using Vivado SDK (C language) and Vivado 2016 (VHDL). This aims to do a comparison between the existing image filtering software such as MATLAB and Field Programmable Gate Array. Through various filtering techniques, I was able to compare and contrast and identify which method is most efficient in processing images as well as which produce a better-quality image. Experiments were run using images of Lena and Barbara, with the resolution of 720 x 1280 to compute the outcomes. Elaborations were done to explain the entire process from converting the images into data outputs, to hexadecimal before it is being sent to Vivado SDK to generate the outcomes. Filtering methods such as grayscale, Sobel filtering and Robert filtering is done via hardware implementation in VHDL code. Then the results generated will also be stored on the zed board and that will be displayed on the monitor through VGA connection. I have also run the exact same images in Matlab to derive the ideal results which are later use to do a comparison during my hardware implementation.