Human Face Recognition Algorithm Based On Wavelet Subbands And Decision Fusion

The technology of face recognition is an active subject in the area of pattern recognition. There are broad applications in the fields of law, business etc. For the particularity of the face image, face recognition is also the very difficult problem. There is still much work to do.Face recognition can be divided into three parts as follows: face detection, feature extraction and classification identification. And the feature extraction of the face images is the most important problem of face recognition. Since 1990, many face recognition algorithms are proposed. We can classify these algorithms into two approaches, namely, constituent and face based. In the constituent-based approach, recognition is based on the relationship between human facial features such as eyes, mouth, nose, profile silhouettes and face boundary. The success of this approach relies highly on the accuracy of the facial feature detection schemes. However, extracting facial features accurately is difficult. Face-based approach (also called as algebraic features-based )attempts to capture and define the face as a whole. The face is treated as a two-dimensional pattern of intensity variation. So, the face-based approach is more attractive.In the face-based approach, the eigenface algorithm, algorithm based on Simgular Value Decomposition and spectroface algorithm are three the most popular and promising algorithms. But, there recognition ratios and recognition speeds needed to be improved. There is still distance to meet the actual application. So many people proposed many algorithms to impove there performance.Wavelet transform(WT) has been a very popular tool -for image analysis in the past ten years. In fact, WT filter the signal using bandpass filter on the different scales. The signal is decomposed into different frequency band, then farther process go on. The advantages of WT, such as good time and frequency localizations, have been discussed in many research articles. WT is applied in face recognition because of the two reasons as follows: 1) By decomposing an image using WT, the resolutions of the subband images are reduced. In turn, the computational complexity will be reduced dramatically by working on a lower resolution image. 2) Wavelet decomposition provides local information in both space domain and frequency domain.After the face images are decomposed by WT into different subbands in suitable levels, not only the latter processing speed is quicken greatly, but also the experiment proved that recognition ratio can be improved. But now many algorithms is only using the subband with thelastest frequency. However, other subbands also can be applied to recognise the face. So in this paper, I propose a face recognition algorithm based on wavelet subbands and decision fusion. We recognise the face image using several subbands. We can get several results for identification. The lastest result is producted according to a dicision fusion method. Both theory and expriment prove this algorithm' s efficiency. Its discriminatory power is higher than the algorighm without using wavelet transform and with single subband.