Research Of Face Recognition Based On Wavelet Transformation

Face recognition is an important technique for human identity recognition when supervising, managing and recognizing etc, and it is also widely used in many other fields. Wavelet Transformation can extracts most essential face features, weaken interferences, reduce complication for it is able to capture both spatial and frequency localization and provide global approximation and local detail information at different scales. So the research of face recognition based on Wavelet Transformation has good foreground.The research work focuses on the application of Wavelet Transform on face recognition. The principal work are listed as follows:1. A new method of facial features localization based on discrete wavelet transformation (DWT) is presented. Our method is more robust for changes of facial pose, expression and covering, while it has fewer computations.2. The face recognition algorithm based on DWT and Principal Component Analysis (PCA) is presented. The based DWT PCA algorithm can get better performance using less computation and less time.3. The face recognition algorithm based on DWT and Artificial Neural Network (ANN) is presented. The algorithm has perfect result. By experiments, The Optimal wavelet base function, wavelet transformation progression and the number of nerve cells in the hidden layer of ANN are determined.4. The face recognition algorithm based on DWT and Support Vector Machine (SVM) is presented. The algorithm can get high recognition rate, even when changes of light, expression, pose or others. As an important part of SVM, The Optimal kernel function and its parameters are determined by experiments.