ECE 734 - Course Project
Sebastian Siegel
Title: Implementation of 2-D FIR filters on systolic arrays for image processing

One major branch in image processing is filtering. It is capable of image
enhancement, edge detection, noise reduction and other useful effects. As a 2-D
FIR filter convolves an image with a filter matrix, the large amount of
computations involved requires a designated hardware to meet real time
constraints. Fortunately, the regularity of the algorithm allows the usage of
systolic arrays.

This project examines existing solutions and compares them to those that will be
derived. The process of mapping an algorithm to a systolic array usually 
involves intermediate steps. After each such step, certain decisions are 
necessary in terms of how to continue. It is one goal of this project to point
out the consequences of each decision made as it may be a constraining force
later on. Furthermore, the issue of symmetry will be addressed if time 
permits.