A multiresolution feature extraction algorithm for face recognition is proposed based on two-dimensional discrete wavelet transform(\r\n2D-DWT), which efficiently exploits the local spatial variations in a face image. For feature extraction, instead of considering\r\nthe entire face image, an entropy-based local band selection criterion is developed, which selects high-informative horizontal\r\nsegments from the face image. In order to capture the local spatial variations within these bands precisely, the horizontal band\r\nis segmented into several small spatial modules. The effect of modularization in terms of the entropy content of the face images\r\nhas been investigated. Dominant wavelet coefficients corresponding to each module residing inside those bands are selected as\r\nfeatures. A histogram-based threshold criterion is proposed to select dominant coefficients, which drastically reduces the feature\r\ndimension and provides high within-class compactness and high between-class separability. The effect of using different mother\r\nwavelets for the purpose of feature extraction has been also investigated. PCA is performed to further reduce the dimensionality\r\nof the feature space. Extensive experimentation is carried out upon standard face databases, and a very high degree of recognition\r\naccuracy is achieved by the proposed method in comparison to those obtained by some of the existing methods
Loading....