Modification of Chaotic Prediction-Based Target Detection Method in the Sea Clutter

The Memory-base nonlinear prediction-based NP-CFAR processing is a typical and advanced method for target detection in the chaotic sea clutter.In consideration of the time-varying instability of the power and character of the sea clutter,authors present a modified method with the rotation hyper-box classification for detection,and use the simpler box-dimension character extraction for preprocessing to reduce the computation load.The simulation experiments prove the efficiency of those modifications.