A Robust and Novel Framework for Subdivision of Face Image Based on the Concept of Golden Rule
Nita M. Thakare · IOSR Journal of Computer Engineering · 2013
To implement any face recognition system, robust face representation and face image normalization is of vital importance.The performance of any face recognition system can greatly be affected by the poor performance of registration technique.3D face recognition methodology has proved as most successful modality.This proposed approach is novel approach and used for subdividing the face into four face-stripes.This technique is based on leo-nardo-da-vinci's golden rule.The 3D face image is represented by modality which represents 2D information using depth map.To develop expression invariant face recognition system the face need to be subdivided into four regions so that the four major parts of face should be separated.The proposed approach had succeeded to present robust results, and insures improved performance of face recognition system.