Using Scale-Space Anisotropic Smoothing for Text Line Extraction in Historical Documents

This paper presents a novel approach for text line extraction which is based on Gaussian scale space, a dedicated binarization, and an energy minimization framework. It enhances the text lines in the image using multi-scale anisotropic second derivative of Gaussian filter bank at the average height of the text line. It then applies a binarization, which is based on component-tree and is tailored for line extraction. The final stage of the algorithm is based on an energy minimization framework for removing spurious text lines and assigning connected components to lines. We developed two algorithms along these ideas, one for documents with mild curled, generally horizontal lines and the other for multiskew and curled text lines. We tested our approach on various datasets written in different languages at varying range of image quality and received high detection rates, which outperform state-of-the-art algorithms.