Wrinkle detection system based on active appearance model

Abstract

The facial wrinkle is a important feature of facial expression, automatic wrinkle detection has been an important topic in recent most works. Many wrinkle detection works focusing on detecting forehead wrinkle line, which are linear shapes, whereas cheek wrinkle lines has not been deeply studied, because of their shape complexity. In this paper presented an effective algorithm for automatic detection of wrinkle line, which largely consist of active appearance model based on wrinkle structure location. The proposed wrinkle detector is applied for real time application to identify early Stroke symptoms. Using unique initial shape instead of active appearance model mean shape, our experimental result got more accurate results than another shape models methods. Experimental results illustrated the competitiveness of the proposed method in detecting wrinkle lines with 92% accuracy.