A classification of emotion and gender using local biorthogonal binary pattern from detailed wavelet coefficient face image

Abstract

This work investigates a framework which identifies gender and emotion of the person from the face image. Gender with their expressions has a vital role in the suspect detection systems. The proposed system aids in identification of a person with their gender as male and female. Also detects gender’s expression as joy and sadness. In this paper, wavelet detailed coefficient and Biorthogonal family-based system have been used simultaneously to identify gender and emotion of a face image. Detailed image local Biorthogonal binary pattern (DILBBP) has been applied for feature extraction and for classification purpose; SVM is applied. Experiments are performed on publicly available standard FERET, INDIAN FACE, and AR FACE databases. Proposed work gives acceptable classification and recognition results with less computational time.