Imerging local feature and fuzzy weighting for 2DPCA face recognition

Two Dimensional Principal Component Analysis (2DPCA) extracts the global feature of human face, but the local feature is very important to face recognition. In this paper , a method of imerging local feature and fuzzy weighting for 2DPCA face recognition was proposed. Firstly , two local feature areas - Upper area and Tzone area were divided, 2DPCA analysis is made to the view picture image and two regions independently. After the initial classify results were fuzzy weighted, minimum distance classification was used for face recognition, the purpose of unifying local feature and overall feature is carried on, the status of eye, nose, mouth and so on local features in the recognition is prominent. The experiments on ORL face databases demonstrate the proposed method's effectiveness and feasibility.