Statistical shape analysis for computer aided facial deformity detection

Abstract

This paper describes craniofacial medical application, with the emphasis on the comparison of normal and abnormal facial data by using statistical shape analysis method to classify the data sets. The data consists of 3D laser scans of mannequin faces. The normal facial dataset is based on the mannequin. The distance between two landmarks and angle between three landmarks are used as the reference to determine whether the facial datasets are normal or abnormal with respect to the normal facial dataset. The statistical shape analysis is a useful tool for facial analysis after the treatment of facial deformity.