Advanced search


Moisés Pastor-i-Gadea, Alejandro H. Toselli, Verónica Romero, Enrique Vidal. Improving Handwritten Off-line Text Slant Correction. Procc. of The Sixth IASTED international Conference on Visualization, Imaging, and Image Processing (VIIP 06), 2006.

Humans can read texts independently of the style they are written. Writing style do not help to identify what is written. For this reason, it is a good idea to make the system invariant to the style. One of the major sources of variability that characterizes the writing style is the slant angle. In this work, a formalization of projection profile methods, is made. New experimental results are obtained using, and consolidating slant correction methods from a previous work on the subject, under this paradigm. Moreover, new projection profile slant correction extensions are proposed and tested.