LBP Feature Extraction for Facial Expression Recognition

摘要

The study of facial feature extraction based on Local Binary Patterns (LBP) is presented in this paper. As a nonparainetric and computationally simple descriptor. LBP is broadly used in facial expression recognition for its invariant to monotonic grayscale transformation. First, the LBP descriptor is presented followed with the facial expression recognition framework using LBP. Aim for the facial expression recognition, the characters of global and local facial feature are analyzed. Then several LBP feature extraction methods with respect to different facial features are described. And the corresponding experiments based on different methods of LBP facial feature extraction are conducted. At last the conclusion is drawn based on the experimental results evaluation.