Face Recognition with Edge Detection and LBP Feature Extraction

The face is the main part of humans that is used as a recognition center in everyday life. With the different faces of many people, it will be increasingly difficult to remember by relying on human memory itself. Therefore we need a method called facial recognition technology. In its application, face recognition consists of two parts, namely face detection and face classification. However, this problem is different from computer recognition, facial recognition is still difficult because of different face shapes. So, in order for the computer to recognize it, a useful method is needed to reduce the complexity of the face. The methods used in face detection are the canny edge detection method and the LBP feature extraction method which will then continue to classify using SOM. This research was conducted with 2 types of datasets, namely full face and eye datasets, which will be converted into vector arrays and will be stored in the training model. From several stages carried out, this research succeeded in obtaining an evaluation accuracy rate of 17% for the full face dataset model and 11% for the eye dataset. The tests carried out were 20 different data, obtained results of 50% accuracy for the full face dataset and 25% for the eye dataset