Night mode face recognition using adaptively weighted sub-pattern PCA

The face recognition problem is made difficult by the great variability in head rotation and tilt, lighting intensity and angle, facial expression, aging, partial occlusion (e.g. Wearing Hats, scarves, glasses etc.), etc. Principal components from the face space are used for face recognition to reduce dimensionality of database images. However, this paper discusses on adaptively weighted sub-pattern PCA (Aw-SpPCA) based face recognition system for dark images that have captured at night. It is really difficult to capture good quality picture at night for lacking of light source with traditional acquisition devices like camera or mobile phone. The computational photographic concepts have been applied to enhance the quality of the capture images at night automatically. Multi-scale retinex color restorations (MSRCR) technique has been applied for overcome this problem. Moreover, for recognition phase of this propose method, unlike PCA based on a whole image pattern, Aw-SpPCA operates directly on its sub patterns partitioned from an original whole pattern and separately extracts features from them. Aw-SpPCA can adaptively compute the contributions of each part and then endows them to a classification task in order to enhance the robustness to both expression and illumination variations. Experimental results show that the proposed method is competitive.