PUNJABI CHARACTER RECOGNITION USING WAVELET AND CURVLET TECHNIQUES WITH SIFT ALGORITHM

There are many techniques used to recognise the characters and out of these one is named as curve let technique and other is wavelet technique. The Punjabi characters has different type of patterns as compared to English characters. Curve let is a non - adaptive technique for multi scale object repr esentation and curved singularities can be well approximated with very few coefficients and remain coherent waveforms under the action of the wave equation in a smooth medium. Standard wave let transforms for two dimensional functions have only very crude capabilities to resolve directional features. The usual orthogonal wave let transforms have wave lets with primarily vertical, primarily horizontal and primarily diagonal orientation. The handwritten characters have different orientations. Therefore the ha ndwritten recognition of characters is challenging research area in the field of pattern recognition. Wavelets generalize the Fourier transform by using both location and spatial frequency. For 2D or 3D signals, directional wavelet transforms go further, b y using basis functions that are also localized in orientation. A curve let transform differs from other directional wavelet transforms in that the degree of localization in orientation varies with scale. In this paper we are using both curve let and wavel et techniques to recognise the Punjabi characters and compare the results.