Heru Susanto
Universitas Muhammadiyah Bengkulu

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Analysis of Face Detection with Head Accessories Using Haar Cascade in Image Processing Heru Susanto; Yuza Reswan; Nuri Veronika; Rozali Toyib
Jurnal Komputer, Informasi dan Teknologi Vol. 5 No. 2 (2025): Desember
Publisher : Penerbit Jurnal Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.53697/jkomitek.v5i2.3146

Abstract

This research is entitled Analysis of Face Detection With Accessories On The Head Using Haar Cascade In Image Processing. The problem faced is how accurate the haar cascade algorithm is in detecting faces with accessories on the head, whether the haar cascade algorithm can still detect faces even though there are accessories that are felt around the face, therefore a study was conducted on face detection with accessories on the head using the haar cascade algorithm in several facial conditions such as facial angles and in several lighting conditions in the detection process. This study aims to analyze the accuracy of the haar cascade algorithm in detecting faces with accessories on the head, whether haar cascade can still detect faces even though there are accessories around the face that block, and 20 facial data sets were taken with a real-time laptop camera. The facial data obtained was taken in several places such as campus areas, houses and boarding houses and with various lighting conditions. From 20 face data taken in real time, researchers obtained 2 face data that were not detected due to poor lighting when taking face data and obtained 18 face data that were successfully detected even though using accessories around the face and it is certain that haar cascade has very good accuracy in detecting faces even though they were blocked by accessories. The testing methods applied are precision, recall and accuracy for calculating the results obtained. The results of the study obtained a precision of 100%, recall 90% and accuracy of 90%.