Facial Asymmetry Analysis Based on 3-D Dynamic Scans

Abstract

Facial dysfunction is a fundamental symptom which often relates to many neurological illnesses, such as stroke, Bell’s palsy, Parkinson’s disease, etc. The current methods for detecting and assessing facial dysfunctions mainly rely on the trained practitioners which have significant limitations as they are often subjective. This paper presents a computer-based methodology of facial asymmetry analysis which aims for automatically detecting facial dysfunctions. The method is based on dynamic 3-D scans of human faces. The preliminary evaluation results testing on facial sequences from Hi4D-ADSIP database suggest that the proposed method is able to assist in the quantification and diagnosis of facial dysfunctions for neurological patients.