VHDL implementation of fuzzy based handwriting recognition system

Abstract

This paper describes the implementation of fuzzy based Handwriting Recognition (HR) system in VHDL. Key motivation of the design is to reduce computational complexity and performance enhancement. Fuzzy intelligent technique is used to tackle inherent ambiguities in different handwriting recognition. This system is capable of recognizing handwritten characters such as one segment, distinguish characters and symbols. The algorithm of the HR system is developed in Matlab first, followed by Fuzzy perception. The model is extensively tested with different sets of handwriting to verify the correct functionality of the system. The Matlab codes are then translated into VHDL using Quartus II from Altera. Simulation results showed clear evidence that fuzzy based HR system is capable of emulating different human reading. Comparison between simulation results in Matlab and VHDL is presented to validate the developed HR system. Quartus Integrated Synthesis (QIS) is used to synthesize the hardware architecture. The system runs at 4.31 MHz with propagation delay of 231 ns. The developed HR hardware system is applicable in automating document-processing task such as mail sorting, check reading in Personal Digital Assistant (PDA), and tablet PCs etc.