A Biometric Solution of Education Environment Using Face Recognition System
Abdellatif Dahmouni, Karim El Moutaouakil, Khalid Satori · International journal of imaging and robotics · 2018
During the last years, several biometric modalities are coming back into the education field. They are mainly employed to automatically control and supervise the school. In this paper, we ensure the previous tasks by using a new face recognition system. Which combines the Local Gradient Probabilistic Pattern (LGPP) with the two Dimensional Wavelet Transform (2DDWT). Firstly, the almost homogeneous areas and the peaks areas are separate according to LGPP confidence interval based on local statistical moments. Secondly, the obtained images are decomposed using 2DDWT in the approximate (LL) sub-band and three details (HL, LH, HH) sub-bands. The approximate sub-band will be used later as reduced face features. Thereafter, the two- dimensional subspace methods, 2DPCA and 2DLDA, are used to extract the final features vector. Finally, Support Vector Machine (SVM) algorithms are used to classify features dataset. The experimental results on ORL and Yale standard databases prove the effectiveness of the proposed system.