Title :
Utilization of Hopfield neural network and quasi-linear model for longshore current pattern simulation from RADARSAT
Author_Institution :
Inst. of Oceanogr., Univ. Coll. Sci. & Technol., Kuala Terengganu, Malaysia
Abstract :
This study introduced a new approach for modeling longshore current pattern from sequential RADARSAT SAR images. Doppler frequency gradient shift from the two sequential RADARSAT images has been estimated. This model utilized to simulate the longshore current pattern. The quasi-linear model used to map the longshore pattern detected by Doppler frequency shift into the real longshore current simulated from significant wave height at breaking zone. A Hopfield neural network was applied to compare between longshore current estimated from Doppler frequency gradient shift and ones modeled from quasi-linear.
Keywords :
Doppler radar; Doppler shift; Hopfield neural nets; oceanographic techniques; radar imaging; remote sensing by radar; synthetic aperture radar; Doppler frequency gradient shift; Hopfield neural network; RADARSAT SAR images; longshore current pattern simulation; quasilinear model; Azimuth; Educational institutions; Frequency estimation; Hopfield neural networks; Marine technology; Neural networks; Neurons; Predictive models; Radar detection; Sea surface;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2003. IGARSS '03. Proceedings. 2003 IEEE International
Print_ISBN :
0-7803-7929-2
DOI :
10.1109/IGARSS.2003.1294552