Please use this identifier to cite or link to this item: https://dspace.upt.ro/xmlui/handle/123456789/7059
Title: Design of floating point arithmetic units and its application in offline signature recognition system embedded on FPGA [articol]
Authors: Ramyarani, Nallathambi
Subbiah, Veerana
Subjects: Field Programmable gate arrays
Floating point-arithmetic
Handwriting recognition
Support Vector Machines (SVM)
Neural Network
Issue Date: 2019
Publisher: Timișoara : Editura Politehnica
Citation: Ramyarani, Nallathambi; Subbiah,Veerana.Design of floating point arithmetic units and its application in offline signature recognition system embedded on FPGA. Timişoara: Editura Politehnica, 2019.
Series/Report no.: Journal of Electrical Engineering;Vol 19 No 2
Abstract: Floating point arithmetic circuits play an important role in scientific computing, signal and image processing applications due to its wide dynamic range and high precision. In this work, floating point Arithmetic and Logic Units (ALUs) architectures are designed, implemented on Field Programmable Gate Arrays (FPGAs) devices and utilized for signature recognition system. Synthesis results proved that log based unit provides faster computations, but increases the area compared with conventional floating point arithmetic units. Hence offline signature recognition system is designed using logarithmic single precision floating point arithmetic units and implemented on FPGA. The person’s signature is classified by Support Vector Machine (SVM) and Neural Network (NN) approach. From the simulation results of the signature recognition system, the person's signature can be identified based on the features. The synthesis results proved that support vector machine classifier occupied only 34% of the FPGA resources available compared with a neural network approach that occupied 68% of the total resources. The various stages of the signature recognition system are analyzed in Xilinx FPGA and MATLAB.
URI: https://dspace.upt.ro/xmlui/handle/123456789/7059
ISSN: 1582-4594
Appears in Collections:Articole științifice/Scientific articles

Files in This Item:
File Description SizeFormat 
BUPT_ART_Ramyarani_f.pdf873.23 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.