Adaptive Extended Local Ternary Pattern (AELTP) for Recognizing Avatar Faces

Many face recognition techniques have been developed during the past decades but the problem remains challenging, especially recognizing non-biological entities or avatars. Local Binary Pattern (LBP) method is one of these techniques which has shown its superiority in recognizing faces. The original LBP operator mainly thresholds pixels in a specific predetermined window based on the gray value of the central pixel of that window. As a result the LBP operator becomes more sensitive to noise especially in near-uniform or flat area regions of an image. To deal with this problem a generalization of the LBP descriptor, Local Ternary Patterns (LTP), came to the presence. In this paper we introduce a new local adapted texture features for efficient avatar face recognition based on the original LTP operator. The proposed technique, Adaptive Extended Local Ternary Pattern (AELTP), shares with the original LTP descriptor being less sensitive to noise. However AELTP is better as it determines the local pattern threshold automatically based on local statistics. Experiments conducted on two virtual world avatar face image datasets show that our technique performs better than original LBP, original LTP and Extended LTP (ELTP) in terms of accuracy.