Automated system for human facial expression recognition (P-0236) (MFN 2979)

Abstract

It is a Computer based facial expression analyzer for facial feature extraction which uses image processing techniques and then a cloned virtual character replicating the facial expressions extracted in the form of an animated expression. Computer based facial expression analysis is implemented by using a database of human facial images with different facial expressions. Mostly communication is effective by facial expression analysis. Humans understand and interpret various facial expressions easily due to natural training from the environment since birth. Neurophysiologic studies have found out that eyebrow, mouth, eyes etc are amongst the most important features for expression recognition. And then according to those facial expressions an algorithm is designed to extract and then replicating those expressions. The one used in this project involves facial image regions and points analysis approach. The analysis is done using various image processing techniques and the expressions found are cloned by using an animated virtual character.