Vessel identification study for non-coherent high-resolution radar

Abstract

This paper presents a vessel identification study based on vessel profile. The study was developed with real data obtained with high-resolution Continuous Wave Lineal Frequency Modulated (CW-LFM) radar. Cases studied in this work are vessels entering and leaving the harbor. Also, in this paper, a comparison between different classification techniques such as Neural Networks, Support Vector Machine and k-Nearest Neighbor is introduced. The differences between normalization methods are evaluated for each classification technique.