Radar correlation imaging method based on convolutional neural network

Abstract

The invention discloses a radar correlation imaging method based on convolutional neural network. The method comprises the following steps of generating a training set and a verification set; designing the architecture and setting parameters of the convolutional neural network; training the convolutional neural network to optimize the weight coefficient; and inputting echo to obtain a high-precision reconstruction result. The radar correlation imaging method based on convolutional neural network provided by the invention performs multi-layer convolution processing on the initial target reconstruction result under the radar correlation imaging system based on the convolutional neural network after training, so as to obtain a high-precision reconstruction result approximating real target information, and the filtering model after training has high computational efficiency and reconstruction precision with no need of repeated iteration optimization, which can effectively solve the problemof the existence of pseudo-scattering points and granular background noise of reconstructed targets in the case of lower SNR and relatively inaccurate reference signal matrix.

Description

technical field [0001] The invention relates to the technical field of radar imaging, in particular to a radar correlation imaging method based on a convolutional neural network. Background technique [0002] With the development of society, radar high-resolution imaging plays an increasingly important role in ensuring national strategic security and promoting national economic development. Among the existing radar imaging technologies, optical radar can perform forward-looking imaging, has short wavelength, high resolution and fast imaging speed, but relies on target radiation, has poor penetration ability to smoke, dust, fog and obstacles, and is vulnerable to Influenced by environmental factors; microwave radar can actively detect and has strong penetrating ability, but due to the low microwave frequency, long wavelength and low angular resolution, and due to the limitation of imaging principle, imaging accumulation time is required, so it is impossible to achieve high fr...

Images