Smart and Portable Attendance System Based on Face Recognition that Implements the Local Binary Pattern (LBP) Algorithm

Abstract

Research about facial recognition-based attendance system implementing artificial intelligence is usually us-ing laptops or personal computers, as a result, the device is not practical because they have large dimensions and heavy, it is not like Fingerprint attendance’s which is widely used nowadays. Therefore, a portable facial recognition-based attendance that applies artificial intelligence are built by means of Raspberry pi 3 B+. The device’s dimensions are 97, 28, and 66 cm for length, width and height respectively. It is work through three stages, the first one is the face detection. At this stage the device captures the user's face, then image preprocessing is carried out by means of Haar cascade. The second stage is the data training by employing local bi-nary pattern (LBP) algorithm, the last stage is the face recognition. The success rate in recognizing face is 60%, 80% and 100% for 1, 3 and 5 training data respectively. The results show that accuracy of the device is getting higher along with the increasing amount of training data.