12727

11651

كارشناسي ارشد

مخابرات سيستم

اصفهان: دانشگاه صنعتي اصفهان، دانشكده برق و كامپيوتر

1396

سيزده، 88ص.: مصور، جدول، نمودار

ص.ع. به فارسي و انگليسي

محمدرضا احمدزاده

سعيد صدري

1396/05/31

كتابنامه

برق و كامپيوتر

مهندسي برق و كامپيوتر

ID11651

Heterogeneous Face Images Recognition by Combination of Visible and Infrared Images Samaneh Shamsian s shamsian@ec iut ac ir Date of Submission 2017 05 27 Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language Farsi Supervisor Dr Mohammad Reza AhmadzadehAbstractLots of researches have been done for face recognition by using visible VIS spectrum images andgreat success has been achieved in this field Despite the achievements face recognition by VISspectrum images in uncontrolled optical conditions is already a challenging issue In such condi tions using different types of imaging specially infrared imaging sensors can help to improve theefficiency of face recognition system in VIS spectrum Recently using near infrared NIR spectrumimages has been considered as an important information source in uncontrolled optical conditions The purpose of this thesis is to improve the efficiency of face recognition system in VIS spectrumby use of NIR spectrum images Following the purpose a method based on deep learning algorithmsnamed coupled convolutional autoencoder CPCAE has been suggested In this method two con volutional autoencoder neural networks CAE have been used for learning compact representationsof NIR and VIS spectrum images As the CAE neural network can only learn compact represen tations of face images related to one domain it can not model the relationships between NIR andVIS images Thus a vector weight has been used to learn a mapping of compact representations ofNIR spectrum images to compact representations of VIS spectrum images The efficiency of thesuggested method on CASIA NIR VIS 2 0 database images has been examined and the accuracyof heterogeneous face images recognition by combination of VIS and NIR images has been studied Laboratory results show that the average accuracy of recognizing NIR spectrum images is 61 52 per cent that shows the advantage of the suggested method to the same method introduced in this field The average accuracy of recognizing VIS spectrum images has been calculated 62 23 percent Keywords 1 Deep learning 2 Heterogeneous face recognition 3 VIS spectrum 4 NIRspectrum 5 Matching

محمدرضا احمدزاده

سعيد صدري