Abstract | ||
---|---|---|
Sea fog can have both negative and positive impacts on humans life. At present, remote sensing has become the main means of long-term and large-scale observation of sea fog. With the improvement of spectral resolution and increase of data volume, the traditional threshold method is simple and convenient as the main method of current sea fog detection, but it's not flexible and accurate enough which causes people need a more automated and intelligent algorithm to achieve efficient sea fog detection. In this article, we use the U-Net deep learning model to construct the sea fog detection model for MODIS multi-spectral images. The main steps include? (1) Data preprocessing, including the PCA method for dimensionality reduction of data; (2) Manual samples extraction with CALIPSO data assist; (3) Construction and training of U-Net sea fog detection model. The experimental results show that the U-Net model can effectively and machine learning method has good potential in sea fog detection. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1109/WHISPERS.2019.8920979 | 2019 10th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS) |
Keywords | Field | DocType |
Sea fog,MODIS,CALIPSO,Detection,Deep Learning | Dimensionality reduction,Computer science,Remote sensing,Data pre-processing,Atmospheric model,Marine layer,Artificial intelligence,Deep learning | Conference |
ISSN | ISBN | Citations |
2158-6268 | 978-1-7281-5295-0 | 0 |
PageRank | References | Authors |
0.34 | 1 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Chunyang Zhu | 1 | 0 | 0.34 |
Jianhua Wang | 2 | 0 | 0.34 |
Shanwei Liu | 3 | 0 | 1.01 |
Hui Sheng | 4 | 0 | 1.35 |
Yanfang Xiao | 5 | 0 | 0.34 |