Abstract—In this paper, we present a novel method to gender classification using a new simple feature extraction which extracts geometric and appearance features at the same time. This feature extraction is performed by computing the derivative in all pixels of face images and then constructing a histogram based on edges magnitudes and directions. The experiments clearly show that the presented method is quite competitive with 95.67% accuracy on FERET database. In addition, the efficiency of the proposed method makes it a good choice for real-time systems which combine face detection and gender classification.
Index Terms—Gender recognition; feature extraction; edges; histogram; FERET database.
Abbas Roayaei Ardakany is with University of Nevada, Reno (e-mail: email@example.com).
Amin Moaven Joula is with the Department of Computer Engineering, Sharif University of Technology, Tehran.
Cite: A. R . A rdakany and A. M. Jou la, "Gender Recognition Based on Edge Histogram," International Journal of Computer Theory and Engineering vol. 4, no. 2, pp. 127-130, 2012.