Target Detection In Sea Clutter Based On Time-frequency Analysis

Sea clutter, referring to the radar backscattering echoes from local oceansurface, makes serious disturbance in target detection and identification capability.So the study on sea clutter has great significance in the design of radar system, radarsignal processing and marine target detection. Currently radar detection technologyunder the background of sea clutter is widely used in maritime safety, maritimerescue, monitoring illegal smuggling and other areas with great military and civilianvalue. Based on the background above, the paper expands the research on measuredsea clutter data collected by high-resolution radar at low grazing angle.Firstly, aiming at the IPIX coherent radar measured sea clutter data, the paperconducts the study of statistical properties by applying some model to fit the realstatistical amplitude model and tests the result in fitting effect goodness; then does aresearch on temporal correlation and de-correlation time of measured sea clutterseries in time domain, Speckle components and Texture components. The resultsshow a statistical behavior comparable with that obtained from experimental data inthe literature.Hilbert-Huang Transform is a new method of time-frequency analysis based onsignal itself, which mainly includes Empirical Mode Decomposition and Hilbertspectral analysis, proposed by a Chinese-American scientist named Huang N E.in1998. Correlation coefficient method is used to remove false modal functioncomponents during EMD decomposition, and a useful method based on EMD torestrain WVD transform cross-interference term of multi-component signal isproposed. Then Hilbert-Huang Transform is applied to analyze and process IPIXcoherent radar measured sea clutter data and achieves weak target detection with therespect of three different aspects under the background of sea clutter: the proportionof the low-frequency IMFs energy, IMFs energy entropy, and roughness of Hilbertspectrum ridge.The paper proposes two useful methods to achieve weak target detection usingIPIX coherent radar measured sea clutter data based on Time-Doppler velocity andlocal backscattering amplitude statistics. The experimental results have indicated that these two methods are valid and can detect target distance unit without anypriori knowledge.For actual surface is complicated, the measured data often contain radarmeasurement noise and dynamic ambient noise from rough sea surface. In order tosuppress the influence on intrinsic characteristics of real sea clutter data by noise,the paper focuses on the threshold de-noising algorithm based on wavelet transform,the simulations show that wavelet threshold processing of measured data can obtainbetter noise reduction effect.Aiming to process IPIX coherent radar measured sea clutter data, the paperdesigns control interface GUI and MATLAB and realizes human-computerinteraction with embedded algorithms.