Yüz tanıma sistemlerinin güvenlik açısından önemi son yıllarda oldukça artmıştır. Çalışmada, Isparta Uygulamalı Bilimler Üniversitesi Teknoloji Fakültesindeki gönüllü öğrenci ve öğretim üyelerinden oluşan 41 kişiye ait ön, yan, üst ve alt yüz görüntüleri alınarak LDA, LBP ve PCA ile SVD yüz tanıma algoritmaları uygulanarak her birine ait model elde edilmiştir. Elde edilen modeller test yüz görüntüleri üzerinde sınıflandırılarak, RMSE ve MAPE performans ölçüt kriterlerine göre değerlendirilerek ön ve yan yüz tanıma sisteminde PCA ve SVD algoritması, üst ve alt yüz tanıma sisteminde ise LBP Algoritmasının en iyi sonucu verdiği tespit edilmiştir.
The importance of face recognition systems in terms of safety has increased significantly in recent years. In the study, the model of each one was obtained by applying the LDA, LBP and PCA face recognition algorithms by taking front, side, top and bottom images of 41 volunteer students and faculty members at the Faculty of Technology of the Isparta University of Applied Sciences. The models obtained were classified on the test face images, evaluated according to the performance standards of RMSE and MAPE, and the PCA and SVD algorithms in the front and side face recognition systems, and the LBP algorithms in the top and bottom face recognition systems were found to have the best results.
The importance of face recognition systems has increased in recent years. In this study, front, side, upper and lower facial images of 41 people consisting of volunteer students and faculty members from Isparta University of Applied Sciences were taken and LDA, LBP-PCA and SVD facial recognition algorithms were applied and their model was obtained. The obtained models were classified on the test face images and evaluated according to the RMSE and MAPE performance criteria. In the front face and side face recognition system, PCA and SVD algorithm, in the upper and lower face recognition system LBP algorithm were found to give the best results.
Alan : Mimarlık, Planlama ve Tasarım; Mühendislik
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|