FACIAL RECOGNITION IDENTIFICATION METHOD USING HAAR FEATURE METHOD

Abstract

The face recognition process aims to identify some of the basic requirements of the system and learn the techniques required for the development of a realtime face recognition system. Facial recognition is one of the pattern recognition approaches for the purpose of identifying a person's face with a biometric approach. In this case Haar Feature is one of the methods to identify face recognition, Haar feature is used to recognize objects based on the simple value of a feature, not the pixel value contained in the object image. In testing this system, it is known that the success rate in the Confusion Matrix calculation on the image is between the actual and predicted values with a Precission value of 93%, meaning that the level of accuracy between the information requested by the user and the answer given by the system is the same. Recall results from the Precission value tend to be inverse to the results of the Recall value. If the Precission result is small then the Recall result is large, because the success rate of the Recall system in finding back a given information is 98%. Precission and Recall results are not the result of Accuracy. The difference is the level of closeness between the predicted results and the actual results of Precission and Recall. Where the actual results are the results found / relevant (True Positive) and the absence of true results / not found (True Negative) and these results are calculated based on the results found. The prediction results that users want, then the result of Accuarcy is 93% which is obtained based on the addition of actual results divided by the prediction results.