As one of the key technologies for Missile-borne SAR matching guidance,matching-suitable areas selec-tion plays an important role for Synthetic aperture Radar(SAR)images matching performance.In this paper,a matching-suitable areas selection method which was from coarse to fine and layer by layer was proposed combined with SAR imaging characteristics.First,the flat terrain areas were found out as the primary matching-suitable areas according to the elevation information.Then,in the primary matching-suitable areas,the regions with rich and sta-ble scattering characteristics were selected as the candidate matching-suitable areas according to the amplitude in-formation.Finally,in the candidate matching-suitable areas,the regions with unique scene features were found out as the final matching-suitable areas on the basis of correlation surface features.In addition,to solve the problem of unstable information quantity feature,the weighted edge strength feature was proposed to measure the stability and information content of the candidate region,in which the edge strength was weighted by the size of the connected region where the pixel in the edge feature image was located.The experiment result showed that the proposed meth-od can select the matching-suitable areas with good matching performance.
synthetic aperture radar(SAR)imagesmatching guidancematching-suitable areas selectionedge strength featurelayer-by-layer screening
NETL
NSTL
万方数据
