Target detection in sea clutter based on ELM

Abstract

Target detection is a hot topic in the research of sea clutter. The solution of this problem can be divided into two aspects. Firstly, find out the different characteristics between the target and sea clutter. Secondly, take advantage of the classifier to realize the feature classification. Thus, we study the characteristics of sea clutter. As a result, the decorrelation time, the K distribution fitting parameters and the Hurst exponent in the FRFT domain are proved to be three feature vectors that can better distinguish the target from sea clutter. Finally, we bring the Extreme Learning Machine (ELM) in the feature classification. Experiment results demonstrate that the chosen feature vectors are effective. Moreover, the ELM is also effective by comparison with SVM.