Multiscale Region Projection Method to Discriminate Between Printed and Handwritten Text on Registration Forms

Abstract

Techniques to identify printed and handwritten text in scanned documents di®er signi¯cantly. In this paper, we address the question of how to discriminate between each type of writing on registration forms. Registration-form documents consist of various type zones, such as printed text, handwriting, table, image, noise, etc., so segmenting the various zones is a challenge. We adopt herein an approach called \ multiscale-region projection" to identify printed text and handwriting. An important aspect of our approach is the use of multiscale techniques to segment document images. A new set of projection features extracted from each zone is also proposed. The classi¯cation rules are mining and are used to discern printed text and table lines from handwritten text. The proposed system was tested on 11 118 samples in two registrationform-image databases. Some possible measures of e±ciency are computed, and in each case the proposed approach performs better than traditional methods.

A multi-feature selection approach for gender identification of handwriting based on kernel mutual information

“…The ICDAR 2013 hassaine competition data set has 475×4 = 1900 handwritten documents. Registration-form documents data set (Tan et al, 2015) has 11118 handwritten documents. Our goal was to identify the gender of the handwriting samples.…”

“…Two real handwritten data sets were used in our experiments for gender prediction which include the ICDAR 2013 sets (Hassaine et al, 2013) and our registration-form document data sets (Tan et al, 2015).…”