A twin prediction method using facial recognition feature

Abstract

The high similarity between identical twin is known to be a great challenge for Face Recognition Technology. As Face Recognition Technology handles identification and the verification of identity claim of a person, it is really important to have a method which can overcome identical twin problem. In this research, we try to demonstrate a model which can predict and compare identical twin. In this method, we used histogram, RGB colors to find the best criteria for matching for initial stage. Later, we used GLCM method, which measures the texture of the images that are used for twin prediction. In the experimental part, we tested our proposed model with different twin dataset and experimental results show very good performance in twin prediction.