Drunk person identification using thermal infrared images

Drunk person identification is carried out using thermal infrared images. The features used for this purpose are simply the pixel values on specific points on the face of the person. It is proved that for a drunk person, the corresponding cluster in the feature space moves far away from its original position for the sober person. The concept behind the proposed approach is based on the physiology-based face identification. For demonstration purposes, Fisher Linear Discriminant approach is used for space dimensionality reduction. The feature space is found to be of very low dimensionality.