Research On Interference Suppression Using Image Segmentation Methods For High Frequency Radars

The over-the-horizon radar(OTHR)plays an important and significant role in early warning,target detection and ocean remote sensing.The OTHR performance could be severely constrained by radio frequency interference(RFI)and sea clutter.Aiming at this problem,this thesis combines signal processing and image processing to propose a new method of interference suppression,which enhances the adaptive processing capability of radar.This thesis combines the theory of signal processing and image processing,mainly including the following:(1)Image generation method and interference feature analysisThe gray-scale image obtained by the gray-scale transformation function is the basis of the research.By analyzing various signals(i.e.sea clutter,radio frequency interference,and target signals)in the image,the image processing method which is suitable for OTHR is determined.(2)Narrowband RFI extraction based on image segmentationNarrowband RFI appears as a straight line in the Range-Doppler Spectrum(RDS),which may be submerged by sea clutter,which may result in inaccurate interference data.Based on the gray image,the Gauss-Laplace operator,edge detection and line detection algorithms are applied to accurately detect the interference region.Then the interference could be recovered and complete suppression of interference is achieved.(3)Wideband RFI suppression based on sea clutter extractionIn the RDS,the wideband RFI is characterized by a stripe shape whose amplitude is smaller than that of the sea clutter.The interference data generally takes a region excluding the Bragg frequency,but the interference suppression is incomplete due to the influence of sea clutter.In this thesis,the constant false alarm algorithm is combined with the image segmentation algorithm of DRLSE to accurately extract the sea clutter region,recover the interference data and suppress the interference.(4)Adaptive Interference Suppression Based on Texture RoughnessIn the OTHR receiving signal,the interference suppression algorithm has no theoretical optimal index as the reference standard,because the parameters of interference and noise are unknown.By analyzing the texture characteristics of non-clutter regions,it is proposed to apply Tamura texture roughness table as the basis for adaptive optimization of suppression algorithm,and optimal suppression performance is achieves by algorithm loop iteration.