MVM-LBP : mean−variance−median based LBP for face recognition

This paper proposes a novel descriptor called Mean-Variance-Median based Local binary pattern (MVM-LBP). The Median (MBP) calculates the difference between median of all pixels in 3 $$\times$$ pixel window and respective eight neighbours. main drawback MBP is that two different windows can have same code which not appropriate face recognition systems. To improve performance system, proposed (MVM-LBP) involves more statistics mean variance uses mean, variance, values to get robust feature vector. check excellence descriptor, method applied on publicly available facial data sets AT&T, faces94. outcomes show gives encouraging results.