Acquisition of Marine Weather Information Suitable for Korean Coastal Area Characteristics Using Deep Learning Techniques

Kim, J.; Kim, J., and Oh, S., 2023. Acquisition of marine weather information suitable for Korean coastal area characteristics using deep learning techniques. In: Lee, J.L.; Lee, H.; Min, B.I.; Chang, J.-I.; Cho, G.T.; Yoon, J.-S., and Lee, J. (eds.), Multidisciplinary Approaches to Coastal and Mari...

Full description

Saved in:
Bibliographic Details
Published in:Journal of coastal research 2024-01, Vol.116 (sp1), p.191-194
Main Authors: Kim, Jiyoon, Kim, Jaemyeong, Oh, Sewoong
Format: Article
Language:English
Subjects:
Algorithms
Artificial intelligence
Coastal management
Coastal research
Coastal zone
Coasts
Deep learning
Eigenvalues
GEO-SPATIAL APPLICATION
Image acquisition
image recognition
Machine learning
marine weather information
maritime safety information
Meteorological data
Pattern recognition
Specificity
Technology
Weather
Winds
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Kim, J.; Kim, J., and Oh, S., 2023. Acquisition of marine weather information suitable for Korean coastal area characteristics using deep learning techniques. In: Lee, J.L.; Lee, H.; Min, B.I.; Chang, J.-I.; Cho, G.T.; Yoon, J.-S., and Lee, J. (eds.), Multidisciplinary Approaches to Coastal and Marine Management. Journal of Coastal Research, Special Issue No. 116, pp. 191-194. Charlotte (North Carolina), ISSN 0749-0208. This study was conducted for the purpose of acquiring weather information from videos acquired on the Korean coast using deep learning technology. Coastal areas have different topographical and meteorological characteristics, so if you want to interpret meteorological phenomena using artificial intelligence, you must build learning data that takes into consideration the specificity of the region and ensure the specificity of the image recognition algorithm. In this study, to support this, the Korean coastal area was designated as a pilot study area, images were acquired, and this was built as artificial intelligence learning data. In the case of coastal areas, it may not be efficient when the deep learning-based image recognition technology acquired in terrain with other topographical characteristics is introduced as it is due to the influence of humidity and strong winds. Therefore, in the image recognition technique, we used f technology to build and recognize pixel patterns and eigenvalues of images, such as binarization, as learning data. As a result of introducing this technique, significant results were obtained in extracting meteorological information from images acquired on the Korean coast.
ISSN:0749-0208
1551-5036
DOI:10.2112/JCR-SI116-039.1