الفهرس | Only 14 pages are availabe for public view |
Abstract The theory and applications of wavelets have undoubtedly dominated the scientific journals in all mathematical, engineering and related fields throughout the last decade. The great capabilities and useful properties of the wavelets have been used and have proven a great success in many applications. In this work, we presented a survey on the theory of the wavelets including the relation between the Fourier transform and the wavelet transform, the multiresolution analysis concept, the wavelet analysis relation to filter banks, orthogonal and biorthogonal wavelet systems, and various examples of wavelet families and their features. The power of wavelet analysis is emphasized by showing its wide range of applications. A survey on character recognition is given with a special attention given to the Arabic case. The Arabic text features are stated and a survey was made on the previous work on Arabic text segmentation and recognition. A character recognition system for the Arabic language writing was proposed. The proposed system consists of two main stages, the segmentation stage and the recognition stage. The segmentation is based on a set of heuristic rules which are driven from the study of the topological and geometrical features of the Arabic characters. The segmentation stage is preceded with a preprocessing stage that is used to normalize the pen width, skew and extract the topological and geometrical features that will be used in the segmentation stage. In the recognition stage, the wavelet transform is used to extract the mutli-resolution features and the extracted wavelet features are used as features in the recognition stage. Using artificial neural networks, the system was trained using a training set and then the system was tested on a test set of Arabic words. The experimental results indicated that the proposed segmentation and recognition systems for Arabic text show high efficiency in segmenting and recognizing Arabic text. Keywords: Wavelets - Preprocessing - Segmentation - Arabic Character Recognition - Image Processing |