Continuous Contour Features of Aerial Target Recognition Techniques Based on ISAR Imaging

Abstract

Since interference noise and limited imaging algorithm etc caused low-quality images, continuous contour information on targets was difficult to extract, and ISAR imaging on aerial targets wasn’t ideal. On this basis, a new method of image processing is proposed to achieve clearer continuous contour features on targets. The method combines denoising technique based on Contourlet transform with improved CV model. First complex ISAR image is transformed into contourlet coefficient by means of contourlet transform. Contourlet coefficient is denoised and complex ISAR image is reconstructed by improved contourlet coefficient. Initial contour on targets is obtained by morphology method and then CV model method is applied to the initial contour, thereby getting better continuous contour features after finite iterations. Eventually, real ISAR echo data are tested, which validates the validity and feasibility of this method.