FACE RECOGNITION MENGGUNAKAN METODE HAAR CASCADE CLASSIFIER DAN LOCAL BINARY PATTERN HISTOGRAM

Abstract

Increasing threats to data security is a consequence of technological advances that continue to develop, especially with the existence of technology that allows data access remotely. It is important to always maintain data security and take preventive measures to prevent data theft. The author proposes the use of facial recognition biometric technology. Face Recognition is one of several biometric technologies that can be used for identity verification systems. The data used amounted to 5 classes, classes with each class having 250 facial images, the total data used amounted to 1250 facial images, consisting of 1000 training data (train) and 250 test data ( tests). The shooting process is done automatically, the system automatically takes 250 faces per class. The next stage is feature extraction using the Local Binary Patterns Histograms (LBPH) method. Furthermore, the system will recognize the detected faces with those in the dataset. The use of the Haar Cascade Classifier and Local Binary Pattern Histogram (LBPH) methods to detect faces using 250 test data obtains an accuracy value of 92%. Testing in real time was carried out 75 times with a distance of 30 cm, 50 cm and 100 cm. The use of the Haar Cascade Classifier and Local Binary Pattern Histogram (LBPH) methods to detect faces in real time obtains an accuracy of 90%.