CN-119471681-B - MIMO multi-view radar associated imaging method, device and medium based on image entropy

CN119471681BCN 119471681 BCN119471681 BCN 119471681BCN-119471681-B

Abstract

The invention discloses an image entropy-based MIMO multi-view radar associated imaging method, device and medium, aiming at a novel problem that the fluctuation of the scattering intensity of a resolution unit along with the uncertainty of a view angle in radar associated imaging causes the relativity mismatch of radar echo and a corresponding imaging resolution unit reference signal, a strategy of 'result guiding' is adopted, from the aspect of inhibiting the energy scattering of the imaging resolution unit, based on sparse reconstruction of an imaging scene, image entropy is added to construct a novel optimization model, and a soft threshold iterative contraction algorithm is adopted to solve the optimization model, so that a high-resolution multi-view MIMO radar associated imaging result is obtained. The invention increases the number of imaging channels, enlarges the observation visual angle, increases the randomness of the radiation field, and can realize high-precision imaging under the condition of fluctuation of the scattering intensity of the resolution unit.

Inventors

CHEN NINGWEI
ZHANG GONG
XIONG QING
XIE JUN
ZHANG QIAN
HE YANSEN

Assignees

南京航空航天大学

Dates

Publication Date: 20260512
Application Date: 20241023

Claims (7)

1. The MIMO multi-view radar associated imaging method based on the image entropy is characterized by comprising the following steps of: (1) The method comprises the steps of obtaining N reference matrixes according to the positions of receiving and transmitting array elements and the positions of imaging units, splicing the N reference matrixes in a time dimension to obtain a dimension-expanding reference matrix A; (2) The method comprises the steps of adopting a Gaussian mixture distribution model to represent fluctuation of scattering intensity of a radar image resolution unit along with the change of a visual angle, receiving N echo signals by N receiving array elements, and splicing the N echo signals in a time dimension to obtain a dimension-expanding echo signal y; (3) On the premise of target sparsity, constructing an MIMO multi-view radar associated imaging model based on image entropy, setting iteration times k=0, and carrying out initial target scattering sparse vector sigma 0 =0; (4) According to an MIMO multi-view radar associated imaging model based on image entropy, carrying out joint processing on an expanded dimension reference matrix A and an expanded dimension echo signal y, adopting a soft threshold iterative contraction algorithm ISTA to solve, and in the (k+1) th iteration, obtaining estimation on a multi-view target scattering coefficient value sigma k+1 based on a solution sigma k of the kth iteration; (5) Setting a maximum iteration number K max and a convergence threshold eta, enabling k=k+1, and judging whether the iteration number is reached or the convergence condition is met If the multi-view radar correlation imaging scattering coefficient value meets the requirement, obtaining the multi-view radar correlation imaging scattering coefficient comprehensive value; if not, returning to the step (4); (6) And converting the obtained target scattering coefficient vector sigma into a two-dimensional imaging area resolution unit scattering coefficient matrix again, and realizing MIMO multi-view radar associated imaging.
2. The image entropy-based MIMO multi-view radar-associated imaging method of claim 1, wherein the implementation procedure of step (1) is as follows: the m-th transmitting array element transmits a signal St m (t), and the n-th receiving array element and the m-th transmitting array element refer to the reference signal of the l-th grid as follows: Wherein, R m and R n are the position vectors of the m-th transmitting array element and the n-th receiving array element respectively, R l is the position vector of the center of the l-th grid, c is the speed of light, t represents the time, and the echo signals received by the n-th receiving array element from the m-th transmitting array element are recorded as follows: Wherein w m,n (t) is the noise signal of the receiving array element, sigma m,n,l is the scattering coefficient corresponding to the center of the first resolution unit under the view angle, the time is discretized, the time sampling number is J, the echo signal and the reference signal received by the nth receiving array element from the mth transmitting array element are time-dimensionally discretized, and the MIMO multi-view radar associated imaging equation is written as follows: y m,n ＝A m,n σ m,n +w m,n Wherein y m,n is an echo vector received by the nth receiving array element from the mth transmitting array element, A m,n is a reference matrix deduced from the view angle, sigma m,n ＝[σ m,n,1 σ m,n,2 … σ m,n,L ] T is a scattering coefficient vector of the view angle, and w m,n is a noise vector; According to the positions of the receiving and transmitting array elements and the positions of the imaging units, N reference matrixes A 1 ,A 2 ,…,A N are obtained, and the N reference matrixes are spliced in the time dimension to obtain an expanded dimension reference matrix A= [ A 1 ,A 2 ,…,A N ] T .
3. The MIMO multi-view radar-associated imaging method based on image entropy according to claim 2, wherein the implementation procedure of step (2) is as follows: The scattering intensity vector sigma l ＝[σ l,1 … σ l,M×N of a certain resolution unit under different view angles obeys a Gaussian mixture distribution model, and the probability density function expression is as follows: wherein K represents the number of Gaussian components, mu i ,∑ i represents the mean and covariance matrix of each Gaussian component, pi i represents the weight coefficient of each Gaussian component; for the nth receive element, the received echo signals are from all M transmit elements: Wherein, sigma n is the comprehensive imaging result of the nth receiving array element, delta sigma m,n ＝σ m,n -σ n represents the difference value between the imaging result corresponding to the mth transmitting channel and sigma n for the nth receiving channel, namely the fluctuation term of the visual angle, A n is the reference matrix corresponding to the nth receiving array element; Echo data of N receiving channels are obtained by adopting a MIMO system, the echo data of multiple receiving channels are spliced in a time dimension by adopting a data layer dimension expansion fusion method, the randomness of a radiation field reference signal is improved by expanding the space dimension, and a dimension expansion imaging equation based on fluctuation of scattering intensity of a resolution unit is deduced by the following formula: Recording device Wherein y is a dimension-expanding echo vector of MIMO radar association imaging, A n is a reference matrix obtained by deduction of an nth receiving array element, diag (A) is a block diagonal matrix taking A 1 ,A 2 ,…,A N as a diagonal element; To take the following measures And delta sigma n ＝σ n -sigma is the difference value between the imaging result sigma n corresponding to the nth receiving channel and the comprehensive inversion imaging result sigma of the MIMO system.
4. The MIMO multi-view radar-associated imaging method based on image entropy as set forth in claim 3, wherein the step (3) is implemented as follows: In the correlated imaging model, a solution is found that makes the number of non-zero elements in the scattering coefficient matrix as small as possible, then the objective function can be written as minimization of the l 0 norm of σ, the objective function in the optimization problem is solved by replacing the l 1 norm, and the model is converted into: In the formula, The i 1 norm optimization method is a linear convex problem; The new definition mode of image entropy under sparse reconstruction system is adopted, the probability defined by gray value occurrence frequency in the traditional image entropy is changed into entropy defined by L p norm form, and for a radar imaging space formed by L imaging grids, the scattering coefficient of the first grid is sigma l , and the corresponding probability is Obtaining Writing the image entropy into an objective function to obtain an imaging model of the MIMO radar correlation imaging under the fluctuation of the scattering intensity of the target resolution unit: Selecting λ 1 and λ 2 ,λ 2 ＝μλ 1 , characterizing the above formula in a more general manner, yields the target expression:
5. the image entropy-based MIMO multi-view radar-associated imaging method of claim 4, wherein the step (4) is implemented as follows: Is provided with In the (k+1) th iteration, based on the solution σ k of the kth iteration, a quadratic approximation is used: Wherein, t= (2λ max (A H A)) -1 ; after adding l 1 norm and image entropy constraint, the optimization model is: Order the For h (|σ l |) at Performing first-order Taylor expansion: According to the principle of ISTA, the final iteration is obtained:
6. an apparatus device comprising a memory and a processor, wherein: a memory for storing a computer program capable of running on the processor; A processor for performing the steps of the image entropy based MIMO multi-view radar correlation imaging method as claimed in any one of claims 1 to 5 when running the computer program.
7. A storage medium having stored thereon a computer program which, when executed by at least one processor, implements the steps of the image entropy based MIMO multi-view radar correlation imaging method of any of claims 1 to 5.

Description

MIMO multi-view radar associated imaging method, device and medium based on image entropy Technical Field The invention belongs to the field of radar-associated imaging signal processing, and particularly relates to an MIMO multi-view radar-associated imaging method, device and medium based on image entropy. Background Radar-related imaging originates from optical ghost imaging, and is a real aperture imaging radar with a new system. The transmitting end transmits a plurality of space-time uncorrelated random modulation signals, a two-dimensional random radiation field is formed in an imaging area, and an image is reconstructed by performing association processing on echo signals and a deduced reference signal matrix. In comparison to conventional real aperture radar imaging, the imaging resolution is related to the randomness of the radiated field, eliminating the need for a bulky antenna array. Meanwhile, the relative rotation between the radar and the target is not limited, and the defect that the relative motion imaging radar depends on motion and is limited by motion is overcome. The radar correlation imaging provides a new thought and direction for solving the defect of the traditional radar imaging, and brings great attention to students at home and abroad. MIMO (multiple input multiple output) multi-view radar associated imaging increases the resolution by increasing the receiving channel, enlarges the observation view angle, but the characteristic that the scattering coefficient of the target fluctuates along with the change of the view angle can also cause interference to radar associated imaging. The existing traditional radar imaging algorithm and radar correlation imaging are mostly based on isotropic ideal point target models. However, radar-related imaging is seriously dependent on the correlation between radar echoes and corresponding imaging resolution unit reference signals, so that the fluctuation of the resolution unit scattering intensity along with the uncertainty of a visual angle in MIMO or MISO (Multiple Input Single Output) radar-related imaging can cause the mismatch of the correlation between radar echoes and corresponding imaging resolution unit reference signals, cause the energy dispersion of imaging resolution units and seriously influence the quality of radar-related imaging. The radar image of the complex target is an estimated value of a scattering distribution function of the target, which is obtained by weighting broadband scattering measurement data of the target, performing coherent integration and the like, and a learner discusses the understanding of the radar image of the complex target and the explanation of pixel values by introducing a basic concept of the scattering distribution function and the scattering distribution function of the target, which are consistent with the definition of a traditional radar scattering section (Radar Cross Section, RCS), and combining the scattering mechanism and radar image analysis of the classical target. Note that the pixel values of the radar image resolution element should not be interpreted directly as RCS levels of the target. The scattering intensity of the radar imaging resolution unit is influenced by the radar radiation frequency, the material of the target, the overall shape and the position of the resolution unit in the whole target, and the uncertainty fluctuation which changes along with the visual angle is displayed. The fluctuation of the scattering intensity of the radar image resolution unit along with the visual angle is different from the fluctuation of the whole RCS of the target along with the visual angle, and cannot be described by a Swerling fluctuation model of the whole RCS of the target along with the time dimension. For this purpose we choose to characterize the fluctuation of the radar image resolution element scattering intensity with the change of the viewing angle by using a non-uniform distribution. At present, a learner provides a sparse total variation regularization algorithm aiming at the influence of scattering intensity fluctuation of a resolution unit on radar associated imaging, and a total variation term is added into a sparse reconstruction model to restrain fluctuation energy. Aiming at the defects of the prior researches, a MIMO multi-view radar associated imaging algorithm based on image entropy is provided, the algorithm adopts a strategy of 'result guiding', and based on imaging scene sparse reconstruction from the aspect of inhibiting imaging resolution unit energy scattering, image entropy is added to construct a new optimization model, and a soft threshold iterative contraction algorithm (ITERATIVE SHRINKAGE-thresholding algorithm, ISTA) method is adopted to solve the optimization model, so that a high-resolution multi-view MIMO radar associated imaging result is obtained. Simulation results show that the invention can realize high-precision imaging under the condition of resolv