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...

Full description

Saved in:
Bibliographic Details
Published in:Transactions on Informatics and Data Science Vol. 2; no. 1; pp. 15 - 26
Main Author: Nidom, Mohammad Saichu
Format: Journal Article
Language:English
Published: Universitas Islam Negeri Profesor Kiai Haji Saifuddin Zuhri Purwokerto 05.03.2025
Subjects:
Adaptive Boosting Algorithm
Haar Cascade Classifier
Haar Like Feature
integral image
ISSN:3064-1772, 3064-1772
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary: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