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
181
3.
AdaBoost training 
In order to classify the image a classifier is created with the help of 
AdaBoost. A classifier is formed by a small number of features from a large 
set of classifier using AdaBoost. Unlike other methods, AdaBoost algorithm 
gives desired region of the object discarding unnecessary background.
4.
Cascading classifiers 
By using a cascade of stages Viola-Jones algorithm eliminates face 
canditates quickly. Elimination happens if the canditate fails to pass the strict 
requirements of that stage. In each stage difficulty increases and if the 
canditate passes all the stages a face is detected. 
Figure 2.
Cascades using for face detection. 
In this work, implementation of Viola-Jones face recognition algorithm 
has been explored in a Raspberry Pi 4 Model B board.It is a single board 
minicomputer with a 1.5 GHz 64-bit quad core ARM Cortex-A72 processor.It 
runs on Linux and requires microSD card to load its operating system since 
it does not have any built-in storage.To get the images from a real-time video 
Pi camera Module 2 is used.Other hardware used in this project includes a 
microSD card with 32GB of storage. For the coding Phyton programming 
language is used. 
References: 
[1] P. Viola and M. Jones, "Rapid object detection using a boosted cascade of simple features," 
Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and 
Pattern Recognition. CVPR 2001, 2001, pp. I-I, doi: 10.1109/CVPR.2001.990517. 
[2] Cen, Kaiqi. “Study of Viola-Jones Real Time Face Detector.” (2016). 
[3] A. Srivastava, S. Mane, A. Shah, N. Shrivastava and B. Thakare, "A survey of face 
detection algorithms," 2017 International Conference on Inventive Systems and Control 
(ICISC), 2017, pp. 1-4, doi: 10.1109/ICISC.2017.8068607. 

---