Publication detail

Comparing Hough transform to other automated methods of feature extraction and matching for document correction

Michal Gavenčiak

Original Title

Comparing Hough transform to other automated methods of feature extraction and matching for document correction

Type

article in a collection out of WoS and Scopus

Language

English

Original Abstract

In some specific cases of cognitive task assessment in children with developmental dysgraphia, experts need to apply advanced modelling and machine learning techniques on a combination of scanned image data and online drawing signals. Nevertheless, in the first step, they usually need to remove external imperfections in the scanned data (such as the rotation of the input image) so that they could be easily registered with a product recorded by a digitizing tablet. The aim of this work is to introduce a method that could be used for such a purpose. We briefly describe and compare a few techniques commonly used for feature point detection. Next, we describe an algorithm based on the Hough transform developed specifically for use with our dataset and compare its effectivity with more generalistic tools. Detected feature points are employed to estimate a homography matrix and to correct the scanned images. The proposed approach was successful in 89\,\% of cases.

Keywords

Developmental Dysgraphia, Hough transform, Feature detection, Handwriting analysis

Authors

Michal Gavenčiak

Released

25. 4. 2023

Publisher

Brno University of Technology, Faculty of Electrical Engineering and Communication

Location

Brno

ISBN

978-80-214-6153-6

Book

Proceedings I of the 29th Student EEICT 2023 Genera papers

Edition

1

Pages from

270

Pages to

274

Pages count

5

URL

https://www.eeict.cz/eeict_download/archiv/sborniky/EEICT_2023_sbornik_1.pdf

BibTex

@inproceedings{BUT184375,
  author="Michal {Gavenčiak}",
  title="Comparing Hough transform to other automated methods of feature extraction and matching for document correction",
  booktitle="Proceedings I of the 29th Student EEICT 2023 Genera papers",
  year="2023",
  series="1",
  pages="5",
  publisher="Brno University of Technology, Faculty of Electrical Engineering and Communication",
  address="Brno",
  isbn="978-80-214-6153-6",
  url="https://www.eeict.cz/eeict_download/archiv/sborniky/EEICT_2023_sbornik_1.pdf"
}