Sea ice detection studies is beneficial for surface climate change analysis, global environmental monitoring and ice hazard prevention. The Global Navigation System-Reflectometer (GNSS-R) technology uses Delay Doppler Map (DDM) in relation to surface type to detect sea ice, and existing studies have made good progress. Feature extraction of DDM data is an important basis to ensure the accuracy of sea ice remote sensing. In this paper, we introduce the convolutional neural network algorithm based on the AlexNet model, and use the northern hemisphere DDM data of TDS-1 satellite in February 2018 for sea ice detection to verify the feasibility of the convolutional neural network algorithm based on the AlexNet model. The experimental results show that the accuracy of sea-ice - seawater oriented detection is 97.71% with a correlation coefficient of 0.89 with the surface data of NOAA station. The AlexNet-based convolutional neural network is confirmed to be used for GNSS-R sea ice detection.
Convolutional Neural Network GNSS-R Sea Ice Detection Based on AlexNet Model
Lect. Notes Electrical Eng.
China Satellite Navigation Conference ; 2022 ; Beijing, China May 22, 2022 - May 25, 2022
China Satellite Navigation Conference (CSNC 2022) Proceedings ; Chapter : 18 ; 190-198
2022-05-07
9 pages
Article/Chapter (Book)
Electronic Resource
English
Convolutional Neural Network GNSS-R Sea Ice Detection Based on AlexNet Model
British Library Conference Proceedings | 2022
|Using Pretrained AlexNet Deep Learning Neural Network for Recognition of Underwater Objects
BASE | 2020
|ALEXNET-BASED INSULATOR SELF-EXPLOSION RECOGNITION METHOD
European Patent Office | 2021
|