Age estimation and gender classification of facial images based on Local Directional Pattern

In order to improve shortage of Local Binary Pattern (LBP) algorithm in the field of facial feature extraction, this paper uses Local Directional Pattern (LDP) algorithm to extract the facial features about age estimation and gender classification. Because gradient information is more stable than gray information, LDP algorithm is not affected easily by noise. After preprocessing images, we use block strategy to represent facial features more fully. And then, we extract features of each block with LDP algorithm, storing the LDP histogram. Finally, we obtain age and gender identification results by Support Vector Machine (SVM). Experimental results on FG-NET face database show that this method has a good effect on age estimation and gender classification.