Application of Wavelet Decomposition to Document Line Segmentation

ACM Computing Classification System (1998): I.7, I.7.5.In this paper an approach to document line segmentation is presented. The algorithm is based on a wavelet transform of the horizontal projective profile of the document image. The projective profile is examined as a one-dimensional discrete signal which is decomposed using the pyramidal wavelet algorithm up to a precise scale, where local minima and maxima are discovered. These local extrema, projected into the input signal, correspond to the spacing between document lines and to the pivots of the lines. The method has been tested on a broad set of printed and handwritten documents and proven to be stable and efficient