Final Year Project v.2

Pose invariant is the major challenges in face recognition application and system. Minimum Average Correlation Energy (MACE) filter is the main ingredient that will be used in this project. MACE filter will be used to recognize face pose variations. The angle of the poses will be taken from -90˚ to +90 ˚. There are other researchers that have done experiment in pose invariant face recognition using other types of techniques such as PCA, Fisher-faces and 3D Linear Subspace. MACE filter also has been used in illumination face recognition. The result for illumination in face recognition using MACE is 100% recognition rate which is better than other filter of face recognition. To undergo pose invariant face recognition there will be two phase which is head pose estimation and face recognition based on pose. The result for the head pose estimation showed 100% when the number of template image is 20 and when the template image is matched with testing image. For different pose, the result shows MACE unable to recognize face. For analysis of face recognition, the performance of filter is low due to the quality of the testing image.