Tezislər / Theses

 
 
 

---

THE 3
rd
 INTERNATIONAL SCIENTIFIC CONFERENCES OF STUDENTS AND YOUNG RESEARCHERS 
dedicated to the 99
th
anniversary of the National Leader of Azerbaijan Heydar Aliyev
171
TURNSTILE ACCESS CONTROL BASED ON FACIAL 
RECOGNITION AND VACCINE PASSPORT VERIFICATION 
Shams Aliyeva
[e-mail]
Baku Higher Oil School
Baku, Azerbaijan 
Supervisor: Ph.D Associate Professor Ali Parsayan
Keywords:
COVID-19, YOLOv3, R-CNN, Face Recognition 
Introduction 
Turnstiles are commonly used at different facilities – they allow only one 
person at a time to pass through and they manage who is allowed to enter. 
Installation of turnstiles allows companies to control entry into their facility 
autonomously and efficiently. Turnstiles are smarter and more reliable than 
they've ever been due to technological developments over the last decades. 
Considering the restrictions due to COVID-19 pandemic, face recognition 
and vaccine passport verification can help touch-free entrance control to be 
safer and easier than ever. There are several methods available for object 
detection. For instance, a sliding window method is used by systems like 
deformable parts models (DPM), in which the classifier is run at regularly 
spaced locations throughout the entire image. Another example can be R-
CNN which uses region proposal method [2]. However, due to the faster 
recognition and performance, object detection is performed with YOLOv3 in 
this project. The main goal of this project is controlling access of academic 
staff, students and personnel into the university building. They are allowed 
to pass through the turnstile only when they have authorization to enter and 
valid vaccine passport. Moreover, according to the information obtained from 
the database, ones with confirmed COVID-19 will not have an entrance 
access.
Algorithm 
YOLO is a novel method for detecting and creating bounding boxes 
around several objects in a picture in real time. The image is only passed 
through the CNN algorithm once to obtain the output, hence the name [1]. 
YOLO algorithm works using the main three techniques. Residual blocks - 
Firstly, the image is divided into various grids. Each grid has a dimension of 
S x S. Second step is bounding box regression. A bounding box is an outline 
that highlights an object in an image. Every bounding box in the image 
consists of width (bw), height (bh), class (c) and bounding box center (bx, 
by). The last step is intersection over union (IOU) which is a phenomenon in 
object detection that describes how boxes overlap. YOLO uses IOU to 
provide an output box that surrounds the objects perfectly [2].

---