Identical twin matching based on multiple feature extractors
Kumbhar Shrikant Sudhakar, P. Nithyanandam · International Journal of Internet Technology and Secured Transactions · 2022
Identical twins have similarities in their face and facial features, due to that, face recognition performance decreases. The proposed system performs identical twin identification with different illumination conditions and different poses. Two methodologies such as grey level co-occurrence matrix (GLCM) and Gabor are used in the proposed system. The first method is the texture based approach, in which the energy, contrast, correlation and homogeneity parameters are calculated by using GLCM which provides the first level of authentication. The second method is the distance-based approach, in which distance between the facial components is detected by using Gabor and thus the second level matching is performed. The GLCM and distance features are used to differentiate the twins. The support vector machine (SVM) is used for the classification. ND-Twin 2009 dataset is used in the proposed system for the testing purpose and the proposed work achieves 90% accuracy.