Comparison of Semantic Segmentation Approaches for Horizon/Sky Line Detection

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F17%3APU122848" target="_blank" >RIV/00216305:26230/17:PU122848 - isvavai.cz</a>

Výsledek na webu

<a href="http://cadik.posvete.cz/papers/ahmad17comparison.pdf" target="_blank" >http://cadik.posvete.cz/papers/ahmad17comparison.pdf</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Comparison of Semantic Segmentation Approaches for Horizon/Sky Line Detection

Popis výsledku v původním jazyce

Horizon or skyline detection plays a vital role towards mountainous visual geo-localization, however most of the recently proposed visual geo-localization approaches rely on user-in-the-loop skyline detection methods. Detecting such a segmenting boundary fully autonomously would definitely be a step forward for these localization approaches. This paper provides a quantitative comparison of four such methods for autonomous horizon/sky line detection on an extensive data set. Specifically, we provide the comparison between four recently proposed segmentation methods; one explicitly targeting the problem of horizon detection, second focused on visual geo-localization but relying on accurate detection of skyline and other two proposed for general semantic segmentation -- Fully Convolutional Networks (FCN) and SegNet. Each of the first two methods is trained on a common training set comprised of about 200 images while models for the third and fourth method are fine tuned for sky segmentation problem through transfer learning using the same data set. Each of the method is tested on an extensive test set (about 3K images) covering various challenging geographical, weather, illumination and seasonal conditions. We report average accuracy and average absolute pixel error for each of the presented formulation.

Název v anglickém jazyce

Comparison of Semantic Segmentation Approaches for Horizon/Sky Line Detection

Popis výsledku anglicky

Horizon or skyline detection plays a vital role towards mountainous visual geo-localization, however most of the recently proposed visual geo-localization approaches rely on user-in-the-loop skyline detection methods. Detecting such a segmenting boundary fully autonomously would definitely be a step forward for these localization approaches. This paper provides a quantitative comparison of four such methods for autonomous horizon/sky line detection on an extensive data set. Specifically, we provide the comparison between four recently proposed segmentation methods; one explicitly targeting the problem of horizon detection, second focused on visual geo-localization but relying on accurate detection of skyline and other two proposed for general semantic segmentation -- Fully Convolutional Networks (FCN) and SegNet. Each of the first two methods is trained on a common training set comprised of about 200 images while models for the third and fourth method are fine tuned for sky segmentation problem through transfer learning using the same data set. Each of the method is tested on an extensive test set (about 3K images) covering various challenging geographical, weather, illumination and seasonal conditions. We report average accuracy and average absolute pixel error for each of the presented formulation.

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

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2017

Kód důvěrnosti údajů

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

Název statě ve sborníku

Proceedings of the International Joint Conference on Neural Networks (IJCNN)

ISBN

978-1-4799-1961-1

ISSN

—

e-ISSN

—

Počet stran výsledku

8

Strana od-do

1-8

Název nakladatele

Institute of Electrical and Electronics Engineers

Místo vydání

Anchorage

Místo konání akce

William A. Egan Civic and Convention Center in A

Datum konání akce

14. 5. 2017

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000426968704091