A geometric invariant scheme for image classification

An effective geometric invariant scheme for shift and scale invariant wavelet feature extraction method for image classification is proposed The feature extraction process involves a normalization followed by an adaptive shift invariant wavelet packet transform. An energy signature is computed for each sub-band of these invariant wavelet coefficients. A reduced subset of energy signatures are selected as feature vector for image classification. Experimental results show that the proposed method can achieve high classification accuracy of 98.5%, and outperforms the other two image classification methods.