Radar forward-looking imaging method for realizing target contour reconstruction

Abstract

The invention provides a radar forward-looking imaging method for realizing target contour reconstruction and belongs to the field of radar forward-looking imaging. For solving the problem that targetcontour edges are not clear in a radar forward-looking imaging process, the method provided by the invention applies a total variation function to the radar forward-looking imaging and reserves target edge information, thereby realizing clear reconstruction of target contours in a forward-looking region while resolution of an azimuth is improved. Based on a convolution model, the problem of improving the resolution of the azimuth is converted into a deconvolution problem; in a deconvolution process, the total variation function is applied to a regularization method, a target function based ongradient constraints is established, weight of edge information in deconvolution is increased, and target edge information is avoided from being lost in the deconvolution process. The method providedby the invention takes a total variation operator as a regularization item for constructing a target function, improves capability of a regularization method in processing of a suddenly changed edgeand avoids a target edge from being split into multiple relatively strong single-point targets and target contour information from being lost.

Description

technical field [0001] The invention belongs to the field of radar forward-looking imaging, in particular to a radar forward-looking imaging method for realizing target contour reconstruction. Background technique [0002] The high-resolution imaging of the radar platform’s forward-looking area is widely used in many fields such as fighter planes’ ground attack, aircraft blind landing navigation, missile precision guidance, sea detection and imaging, target tracking and identification, and battlefield search. The imaging results with a clear outline can provide more useful information about the target, which is more conducive to target recognition and detection. [0003] Due to the limitations of the imaging mechanism, the traditional synthetic aperture technology and Doppler beam sharpening technology cannot achieve high-resolution imaging in the forward-looking area; while the real beam scanning radar, due to the limitation of the real aperture size, the azimuth resolution...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem that the edge of the target contour is not clear during the radar forward-looking imaging process, the present invention proposes a radar forward-looking imaging method to realize the reconstruction of the target contour. In the convolution inversion process, the total variogram is used to add gradient constraints to Preserve the target edge information and improve the ability of the convolution inversion method to deal with abrupt boundaries, so as to obtain clear target contour reconstruction results

Images