Study On Fractal Method For Maritime Target Detection

Sea clutter refers to the backscattered returns from a patch of the sea surfaceilluminated by a radar pulse. Accurate understanding and characterizing modeling ofsea clutter is crucial in remote sensing and targets detection within sea clutter.Recently, the nonlinear theory has been introduced into many fields, includingradar signal processing. Based on previous researches, this thesis primarily discussesthe application of fractal theory in the modeling of sea clutter and maritime targetdetection.First, this thesis presents the study on the fractal properties of sea clutter,especially the estimation of the box-counting dimension and multifractal spectrum.Second, an integral test method, which was based on the difference betweenmultifractal spectra of sea clutter with and without targets, is devised to detect targets.Third, this thesis analyzes the main performances of fractal methods throughtheoretical analysis and experimental verification. The relationship between thebox-counting dimension and the SCR, the relationship between the integral test and theSCR, and the change of fractal features with the length of utilized series are included.It is shown that the integral test method has a stronger performance of targetdetection than traditional moving target detection. It should be pointed out that theconclusions in this thesis are staged results of our work and need further research.