FPGA Implementation of Various Division Algorithms for Image Processing Applications - A Comparative Analysis

Division is one of the most commonly sort after algorithm for performing image processing operations such as normalization, filtering, enhancement, deconvolution etc. Hence, the design of efficient division algorithm is highly essential in order to obtain a better image processing algorithm The first and foremost step in the implementation of any algorithm in hardware is to identify the number systems used to represent the inputs and outputs. The inputs, outputs and complexity of the algorithm varies based on the number system used. This paper proposes a recursive subtraction based fixed-point division algorithm. This work also provides a comparative analysis on various aspects such as area, power, delay and throughput, on comparing the proposed recursive subtraction based fixed-point division algorithm with the existing floating-point division, restoring integer division and non-restoring integer division algorithms and operator based fixed-point division algorithm.