Research On Weak Target Detection Algorithm In Sea-Clutter Environment

The sea clutter low RCS weak target detection in the field of military and civilianareas has important applications, lead to a higher probability of false alarm due to thesea clutter like the target characteristic as well as the weak target in low signal-to-noiseratio (SCR),The target can not effectively be detected in the clutter area. Traditional seaclutter in the context of target detection algorithm is developed based on the magnitudeof the sea clutter statistical model detection strategy. However, with the continuousimprovement of modern radar range resolution, in the case of low grazing angle seaclutter returns average RCS reduced, sea clutter echo amplitude model graduallydeviated from the Gaussian statistical model, thus leading to the sharp decline inperformance of the conventional detector. Newly discovered sea clutter model is limitedby many conditions, to adapt to the relatively narrow scope, so seek new Sea Cluttersmall target detection algorithm has become a hot research topic. In this thesis, novel seaclutter low RCS target detection methods are studied. The main contributions of thisthesis are as follows:1.First, the formation mechanism of sea clutter is analyzed as well as thecomposition of the sea, to clarify the main components of sea clutter is constituted bythe gravity waves caused by the wind storm and surface tension induced capillary waves.Sea clutter distribution model is analyzed and the the IPIX radar measured data of theamplitude is statistical analysised and fitting, fitting contrast with the commonly usedstatistical distribution model of sea clutter amplitude, and fitting results have been tested.Simulation results demonstrate the best fit K distribution model in this group of seaclutter data.2．K independent and identically distributed joint probability density function of thesecondary data derived fixed-point covariance matrix for the initial matrix use Toeplizematrix T-FP method. The method was measured in non-Gaussian clutter to test thedetection performance. Simulation results show that compared with several commonlyused clutter covariance matrix structure, T-FP method in the case of relatively smallsignal-to-clutter can be obtained better detection results.3.High-speed small target detection algorithm based on Radon transform in strongsea clutter environment is proposed, the algorithm take Radon transform onhigh-resolution radar echo data into parameter field. In parameter field the targetpresented linear characteristics is converted into a peak intensity, threshold and theclustering process is then carried out for each peak, to detect the target in energy and angle domain, the energy accumulated in the transformation process, the sea clutter issuppressed, target can be detected in low SCR conditions. Then on the basis of theRadon transform, the Radon inversion transform weak target detection algorithm isproposed, the methods can suppress the clutter, the target can be transformed back to thetime-range domain, the weak target track information can be clearly detected.4.Multi-fractal characteristics of the real sea clutter data are analyzed. Weresearched the difference between the no-target pure sea clutter with sea clutter mixedtarget in. Proposed a mean square test methods based multi-fractal characteristic oftarget. Construct the new sea clutter data series with different SCR to test the proposedmethod performance. The thesis analyzed the fractal characteristics of the sea clutterand relationship of target and sea clutter, also analyzed the relationship of boxdimension and SCR, relationship of mean square test and SCR. Simulation results showthe algorithm can detect the targets in low RCS.