<![CDATA[The principle of facial recognition is that the facial object captured by the camera will be processed and compared with all the images of the face in the existing data set, so that the identity of the face is known. One of the applications of facial recognition is to conduct attendance with individual faces. In this study, a system was created that can detect and recognize a person's face which is using the Local Binary Pattern Histogtram (LBPH) method. The programming languages used are Python, openCV and Numpy modules. The Javascript programming language is used for the user interface, so the user scans the face through a browser with a MySQL database to store the identity data and name of the owner of the face. The results of the study using the LBPH face method were successfully identified and the data was saved to the database used for attendance data. The amount of training imagery data and the distance of the object to the camera as well as the quality of the camera resolution affect the results of facial recognition so several tests were carried out. Too far away about 120 cm and above the camera, the face cannot be recognized well because the system finds it difficult to capture the pixel area of facial features. The large number of training images for each test image is also quite influential, the more training images for each test image, the better the percentage of recognition success and vice versa. Furthermore, testing with different camera resolution qualities, the higher the resolution of the camera used, the accuracy of detection and recognition becomes accurate and faster. Keywords –OpenCV, Numpy, Facial Recognition, LBPH Method]]>
