Haar Cascade Classifier and Adaboost Algorithm for Face Detection with the Viola-Jones Method

Face detection is a significant challenge in image processing and computer vision, with broad security, identity recognition, and human-computer interaction applications. This study explores the effectiveness of the Haar Cascade Classifier method optimized with Adaboost to improve the accuracy and e...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	Transactions on Informatics and Data Science Ročník 2; číslo 1; s. 15 - 26
Hlavný autor:	Nidom, Mohammad Saichu
Médium:	Journal Article
Jazyk:	English
Vydavateľské údaje:	Universitas Islam Negeri Profesor Kiai Haji Saifuddin Zuhri Purwokerto 05.03.2025
Predmet:	Adaptive Boosting Algorithm Haar Cascade Classifier Haar Like Feature integral image
ISSN:	3064-1772, 3064-1772
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Popis
Shrnutí:	Face detection is a significant challenge in image processing and computer vision, with broad security, identity recognition, and human-computer interaction applications. This study explores the effectiveness of the Haar Cascade Classifier method optimized with Adaboost to improve the accuracy and efficiency of face detection in various head covering conditions. In this experiment, two approaches were compared: using the Haar Cascade Classifier independently and in combination with Adaboost, with evaluation based on metrics such as accuracy, precision, sensitivity, and F1-Score. The results showed that the Adaboost combination significantly improved detection accuracy, with the "Hooded" class achieving an accuracy of 99.2% and the average detection time reduced from 14.9 seconds to 1.9 seconds. These findings show that the use of optimization techniques such as Adaboost not only improves detection performance but also overall system efficiency. The conclusion of this study emphasizes the importance of combining methods in developing a more robust and efficient face detection system. The implications of this research can be applied to create more effective security and facial recognition applications and pave the way for further study in optimizing object detection algorithms.
ISSN:	3064-1772 3064-1772
DOI:	10.24090/tids.v2i1.12276