Towards Historical Map Analysis using Deep Learning Techniques

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F23%3A43969452" target="_blank" >RIV/49777513:23520/23:43969452 - isvavai.cz</a>

Výsledek na webu

<a href="https://link.springer.com/chapter/10.1007/978-3-031-34111-3_16" target="_blank" >https://link.springer.com/chapter/10.1007/978-3-031-34111-3_16</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/978-3-031-34111-3_16" target="_blank" >10.1007/978-3-031-34111-3_16</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Towards Historical Map Analysis using Deep Learning Techniques

Popis výsledku v původním jazyce

This paper presents methods for automatic analysis of historical cadastral maps. Our goal is to detect important features in individual map sheets to allow their further processing and connecting the sheets into one seamless map that can bebetter presented online. We concentrate on detection of the map frame,which defines the important segment of the map sheet. Other crucial features are so-called inches that define the measuring scale of the map. We also detect the actual map area. We propose novel segmentation approaches that combine standard computer vision techniques with neural nets (NNs). We have shown that combining the standard computer vision techniques with NNs can outperform the state-of-the-art approaches in the scenario when only little training data is available. We have also created a novel annotated dataset that is used for network training and evaluation.

Název v anglickém jazyce

Towards Historical Map Analysis using Deep Learning Techniques

Popis výsledku anglicky

This paper presents methods for automatic analysis of historical cadastral maps. Our goal is to detect important features in individual map sheets to allow their further processing and connecting the sheets into one seamless map that can bebetter presented online. We concentrate on detection of the map frame,which defines the important segment of the map sheet. Other crucial features are so-called inches that define the measuring scale of the map. We also detect the actual map area. We propose novel segmentation approaches that combine standard computer vision techniques with neural nets (NNs). We have shown that combining the standard computer vision techniques with NNs can outperform the state-of-the-art approaches in the scenario when only little training data is available. We have also created a novel annotated dataset that is used for network training and evaluation.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2023

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

AIAI 2023: Artificial Intelligence Applications and Innovations

ISBN

978-3-031-34110-6

ISSN

1868-4238

e-ISSN

1868-422X

Počet stran výsledku

13

Strana od-do

173-185

Název nakladatele

Springer Nature Switzerland AG

Místo vydání

Cham

Místo konání akce

León, Španělsko

Datum konání akce

14. 6. 2023

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—