Back to Search Start Over

Coastal Visibility Distance Estimation Using Dark Channel Prior and Distance Map Under Sea-Fog: Korean Peninsula Case

Authors :
Tae Wuk Bae
Jin Hyun Han
Kuk Jin Kim
Young Taeg Kim
Source :
Sensors, Vol 19, Iss 20, p 4432 (2019)
Publication Year :
2019
Publisher :
MDPI AG, 2019.

Abstract

Commercial visibility sensors among meteorological sensors estimate the visibility distance based on transmission, backward scattering, and forward scattering principle. These optical visibility sensors yield comparatively accurate local visibility distance. However, it is still difficult to obtain comprehensive visibility information for a wide area, such as the coast or harbor due to the sensor structure using straightness and scattering properties of light. In this paper, we propose a novel visibility distance estimation method using dark channel prior (DCP) and distance map based on a camera image. The proposed method improves the local limit of optical visibility sensor and detects the visibility distance of a wide area more precisely. First, the dark channel for an input sea-fog image is calculated. The binary transmission image is obtained by applying a threshold to the estimated transmission from the dark channel. Then, the sum of the distance values of pixels, corresponding to the sea-fog boundary, is averaged, in order to derive the visibility distance. This paper also proposes a novel air-light and transmission estimation technique in order to extract the visibility distance for an abnormal sea-fog image, including any light source, such as sunlight, reflection light, and illumination light, etc. The estimated visibility distance was compared with optical visibility distance of an optical visibility sensor and their agreement was evaluated.

Details

Language :
English
ISSN :
14248220
Volume :
19
Issue :
20
Database :
Directory of Open Access Journals
Journal :
Sensors
Publication Type :
Academic Journal
Accession number :
edsdoj.77ee6c40bd3442eab2b1aa9b37581915
Document Type :
article
Full Text :
https://doi.org/10.3390/s19204432