Adaptive single-bit quantization SAR imaging method and device, storage medium and product

Abstract

Embodiments of the present disclosure disclose a self-adaptive single-bit quantization SAR imaging method and device, a storage medium and a product. The method comprises: iteratively reconstructing an image matrix, in each iteration, transforming the image matrix to an echo domain based on a threshold matrix, an original echo signal matrix and an inverse imaging operator, performing single-bit quantization on the image matrix in the echo domain to obtain an echo residual; performing matched filtering on the echo residual using an imaging operator and then performing gradient descent to obtain a gradient descent result matrix; performing threshold shrinkage on the gradient descent result matrix based on a weight matrix and a fractional order sparse norm to obtain an updated image matrix; updating the weight matrix based on the updated image matrix; updating the threshold matrix using a delta modulation algorithm; and obtaining a SAR imaging result when a maximum number of iterations is reached. The method can compensate for the loss of amplitude information and improve the SAR imaging accuracy by adaptively updating the threshold and the fractional order sparse norm constraint while only retaining the signs of the echo signals.

Description

Technical Field

[0001] This disclosure relates to the fields of radar signal processing and compressed sensing imaging technology, and in particular to an adaptive single-bit quantization SAR imaging method, apparatus, storage medium and product. Background Technology

[0002] Synthetic Aperture Radar (SAR) possesses all-weather, all-day imaging capabilities and has wide applications in remote sensing mapping, vegetation monitoring, and military reconnaissance. Traditional SAR imaging systems are limited by the Nyquist sampling theorem, requiring high-precision, high-sampling-rate analog-to-digital converters (ADCs), resulting in high hardware costs and massive data volumes. The introduction of compressed sensing theory reduces sampling pressure, while single-bit compressed sensing further quantizes measurements into binary data containing only symbolic information, allowing the use of low-cost single-bit ADCs and significantly reducing system complexity.

[0003] However, existing single-bit compressed sensing SAR imaging methods quantize data by comparing the measured signal with a zero threshold. Therefore, they do not rely on the absolute amplitude information of the signal, but only retain the relative amplitude information, which leads to the loss of amplitude information of the measured signal, thus limiting the imaging accuracy. Summary of the Invention

[0004] In view of this, embodiments of the present disclosure provide an adaptive single-bit quantization SAR imaging method, apparatus, storage medium, and product, which can significantly improve the accuracy of single-bit SAR imaging by adaptively updating thresholds and fractional-order sparse norm constraints while retaining only the sign of the echo signal, thereby compensating for the loss of amplitude information.

[0005] In a first aspect, embodiments of this disclosure provide an adaptive single-bit quantization SAR imaging method, employing the following technical solution: Obtain the original echo signal matrix, imaging operator, and inverse imaging operator, and initialize the image matrix, threshold matrix, and weight matrix; The image matrix is ​​reconstructed iteratively. In each iteration, based on the threshold matrix, the original echo signal matrix, and the inverse imaging operator, the image matrix is ​​transformed to the echo domain and then quantized by a single bit to obtain the echo residual. The imaging operator is used to perform matched filtering on the echo residual to obtain the filtering gradient, and gradient descent is performed on the filtering gradient to obtain the gradient descent result matrix. Based on the weight matrix and the fractional sparse norm, the gradient descent result matrix is ​​thresholded and shrunk to obtain the updated image matrix; The weight matrix is ​​updated based on the updated image matrix; The threshold matrix is ​​updated using the delta modulation algorithm; Based on the updated threshold matrix and weight matrix, the next iteration is performed. When the current iteration count reaches the maximum iteration count, the contracted matrix obtained from the last iteration is used as the SAR imaging result.

[0006] Optionally, the step of thresholding the gradient descent result matrix based on the weight matrix and the fractional-order sparse norm to obtain the updated image matrix includes: The gradient descent result matrix is ​​extracted row by row and its absolute value is taken to obtain the magnitude vector of the current row; The weight matrix is ​​extracted row by row to obtain the weight vector of the current row; Based on the fractional sparse norm, the product of the weight vector, regularization parameter and iteration step size of the current row is calculated element-wise to obtain the adaptive threshold vector of the current row. Based on the adaptive threshold vector of the current row, threshold shrinkage is performed on the magnitude vector of the current row; After threshold shrinking of all rows of the gradient descent result matrix, the updated image matrix is ​​obtained.

[0007] Optionally, the threshold shrinking of the magnitude vector of the current row based on the adaptive threshold vector of the current row includes: Compare each element in the magnitude vector of the current row with the threshold at the corresponding position in the adaptive threshold vector of the current row; Mark elements that are greater than the corresponding threshold as valid elements, and set the rest of the elements to zero; Based on the fractional sparse norm, the shrinkage coefficient is obtained; The effective element is subjected to amplitude shrinkage based on the shrinkage coefficient; The effective elements and zero-value elements after amplitude contraction constitute the amplitude vector of the current row after contraction.

[0008] Optionally, updating the weight matrix based on the updated image matrix includes: The updated image matrix is ​​extracted element by element and its absolute value is taken. The absolute value of the extracted element is added to a preset constant and the reciprocal is taken to obtain the updated element. Arrange all updated elements in their original positions to form the updated weight matrix.

[0009] Optionally, updating the threshold matrix using the δ modulation algorithm includes: In the first iteration, the simulated echo signal matrix obtained after transforming the image matrix to the echo domain is separated into a real part signal matrix and an imaginary part signal matrix; Obtain the sign deviation between the real part signal matrix and the imaginary part signal matrix and the current threshold matrix, and update the threshold matrix according to the sign deviation and the initial quantization step size; In non-first iterations, the quantization step size is updated based on the consistency of the sign deviation between the two iterations. The threshold matrix is ​​updated based on the updated quantization step size and the current sign deviation.

[0010] Optionally, the sign bias includes real part sign bias and imaginary part sign bias, and the quantization step size includes real part quantization step size and imaginary part quantization step size; updating the quantization step size based on the consistency of the sign biases of two consecutive steps includes: If the current imaginary part sign deviation is the same as the previous real part sign deviation, then increase the real part quantization step size; If the current imaginary part sign deviation is different from the previous real part sign deviation, then the real part quantization step size is reduced; If the current imaginary part sign deviation is the same as the previous imaginary part sign deviation, then increase the imaginary part quantization step size; If the current imaginary part sign deviation is different from the previous imaginary part sign deviation, then the imaginary part quantization step size is reduced.

[0011] Optionally, the fractional sparse norm is the L2 / 3 norm.

[0012] Secondly, this disclosure also provides an adaptive single-bit quantization SAR imaging system, which adopts the following technical solution: The acquisition module is used to acquire the original echo signal matrix, imaging operator and inverse imaging operator, and initialize the image matrix, threshold matrix and weight matrix; An iterative module is used to iteratively reconstruct the image matrix. In each iteration, based on the threshold matrix, the original echo signal matrix, and the inverse imaging operator, the image matrix is ​​transformed to the echo domain and then quantized by a single bit to obtain the echo residual. The filtering module is used to perform matched filtering on the echo residual using the imaging operator to obtain the filtering gradient, and to perform gradient descent on the filtering gradient to obtain the gradient descent result matrix. The shrinkage module is used to perform threshold shrinkage on the gradient descent result matrix based on the weight matrix and the fractional sparse norm to obtain an updated image matrix; The first update module is used to update the weight matrix based on the updated image matrix; The second update module is used to update the threshold matrix using a delta modulation algorithm; The imaging module is used to perform the next iteration based on the updated threshold matrix and weight matrix. When the current iteration count reaches the maximum iteration count, the shrunken matrix obtained from the last iteration is used as the SAR imaging result.

[0013] Thirdly, this disclosure also provides a computer device, which adopts the following technical solution: The computer device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform any of the adaptive single-bit quantization SAR imaging methods described above.

[0014] Fourthly, embodiments of this disclosure also provide a computer-readable storage medium storing computer instructions for causing a computer to execute any of the adaptive single-bit quantization SAR imaging methods described above.

[0015] Fifthly, embodiments of this disclosure also provide a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of any of the methods described above.

[0016] The adaptive single-bit quantization SAR imaging method disclosed in this embodiment transforms the image matrix to the echo domain and then performs single-bit quantization, retaining the advantages of low hardware cost and low data volume. It introduces a delta modulation algorithm to dynamically update the threshold matrix, allowing the quantization threshold to adaptively adjust according to signal changes, thereby approximating the amplitude of the true echo signal. This method effectively compensates for the defect of completely discarding amplitude information in traditional zero-threshold quantization, laying the foundation for subsequent high-precision reconstruction. In image reconstruction, this scheme employs threshold shrinkage based on the fractional-order sparse norm and dynamically updates the weight matrix according to the updated image matrix, enabling the sparsity constraint to adaptively enhance the target and suppress noise and artifacts during the iteration process. Compared with traditional regularization methods, the fractional-order sparse norm has a stronger sparse approximation capability, thus significantly improving the detail preservation and noise robustness of the reconstructed image while retaining the advantages of single-bit quantization. Finally, the high-resolution SAR image obtained through iterative optimization convergence effectively solves the problem of decreased imaging accuracy caused by the loss of amplitude information.

[0017] The above description is merely an overview of the technical solution disclosed herein. In order to better understand the technical means of this disclosure and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating the adaptive single-bit quantization SAR imaging method provided in this embodiment of the disclosure; Figure 2 A flowchart illustrating the threshold shrinkage method for the gradient descent result matrix provided in this embodiment of the disclosure; Figure 3 A flowchart illustrating the threshold shrinkage method for amplitude vectors provided in this embodiment of the disclosure; Figure 4 A flowchart illustrating the weight matrix update method provided in this embodiment of the disclosure; Figure 5 A flowchart illustrating the threshold matrix update method provided in this embodiment of the disclosure; Figure 6 The image obtained by imaging the Chirp Scaling algorithm provided in the embodiments of this disclosure; Figure 7 The image obtained by imaging using the L1-IST algorithm provided in this embodiment of the disclosure; Figure 8 The image shows the result of imaging using the L1 / 2-IST algorithm provided in this embodiment of the disclosure; Figure 9 The image shows the imaging result of the adaptive single-bit quantization SAR imaging method provided in the embodiments of this disclosure. Figure 10 This is a block diagram of the adaptive single-bit quantization SAR imaging system provided in the embodiments of this disclosure; Figure 11 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present disclosure. Detailed Implementation

[0020] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0021] It should be understood that the following specific examples illustrate the implementation of this disclosure, and those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0022] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0023] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this disclosure. The drawings only show the components related to this disclosure and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0024] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0025] Reference Figure 1 This disclosure provides an adaptive single-bit quantization SAR imaging method, comprising the following steps: S1: Obtain the original echo signal matrix, imaging operator, and inverse imaging operator, and initialize the image matrix, threshold matrix, and weight matrix; S2: Iteratively reconstruct the image matrix. In each iteration, based on the threshold matrix, the original echo signal matrix, and the inverse imaging operator, the image matrix is ​​transformed to the echo domain and then quantized by a single bit to obtain the echo residual. S3: Use the imaging operator to perform matched filtering on the echo residual to obtain the filtering gradient, and perform gradient descent on the filtering gradient to obtain the gradient descent result matrix; S4: Based on the weight matrix and the fractional sparse norm, the gradient descent result matrix is ​​thresholded and shrunk to obtain the updated image matrix; S5: Update the weight matrix based on the updated image matrix; S6: Update the threshold matrix using the δ modulation algorithm; S7: Based on the updated threshold matrix and weight matrix, perform the next iteration. When the current iteration count reaches the maximum iteration count, use the shrunken matrix obtained from the last iteration as the SAR imaging result.

[0026] The adaptive single-bit quantization SAR imaging method disclosed herein transforms the image matrix to the echo domain and then performs single-bit quantization, retaining the advantages of low hardware cost and low data volume. It introduces a delta modulation algorithm to dynamically update the threshold matrix, enabling the quantization threshold to adaptively adjust according to signal changes, thereby approximating the amplitude of the true echo signal. This method effectively overcomes the defect of completely discarding amplitude information in traditional zero-threshold quantization, laying the foundation for subsequent high-precision reconstruction.

[0027] This scheme employs threshold shrinkage based on fractional-order sparse norm in image reconstruction and dynamically updates the weight matrix according to the updated image matrix. This allows the sparsity constraint to adaptively enhance the target and suppress noise and artifacts during the iteration process. Compared with traditional regularization methods, fractional-order sparse norm has stronger sparse approximation capabilities, thus significantly improving the detail preservation and noise robustness of the reconstructed image while retaining the advantages of single-bit quantization. Finally, the high-resolution SAR image obtained through iterative optimization effectively solves the problem of decreased imaging accuracy caused by the loss of amplitude information.

[0028] In S1, the original echo signal matrix is ​​acquired by the radar receiver, the imaging operator and its inverse operator are pre-constructed based on the radar system parameters, and the image matrix is ​​the SAR image reconstructed in the current iteration and is initially set to a zero matrix (0). m×n The threshold matrix is ​​used for single-bit quantization comparison and is initially set randomly, while the weight matrix is ​​used for weighted sparsity constraints and is initially set as an identity matrix.

[0029] In S2, the maximum number of iterations, the current number of iterations, and the iteration step size are also initialized. The image matrix is ​​reconstructed iteratively according to the iteration step size. Each iteration increments the current number of iterations by one until it equals the initialized maximum number of iterations, at which point the iteration stops.

[0030] In each iteration, the inverse imaging operator transforms the current image matrix to the echo domain, generating a simulated echo signal matrix. A threshold matrix is ​​then used to perform single-bit quantization on the simulated echo signal matrix; this involves calculating the difference between the simulated echo and the threshold, taking its sign, and obtaining the single-bit quantization result. The echo residual is obtained by subtracting the single-bit quantization result from the original echo signal matrix. The formula for calculating the echo residual is as follows: In the formula, Indicates the current iteration number; Indicates the first The echo residual after the nth iteration; Represents the original echo signal matrix; This function indicates whether a number is positive or negative; it outputs +1 for positive numbers and -1 for negative numbers. Represents the inverse imaging operator; Indicates the first Image matrix after n iterations; Indicates the first Transform the image matrix to the echo domain after several iterations; Indicates the first The threshold matrix for the number of iterations; This indicates that the analog echo signal matrix is ​​quantized using a single bit.

[0031] In S3, the imaging operator transforms the echo residual to the image domain, which is essentially a matched filtering operation. The result is the filtered gradient (also called the gradient direction or update direction) in gradient descent. The filtered gradient is scaled by the iteration step size, and the scaled gradient components are summed with the current image matrix to complete one round of gradient descent iteration. The formula for calculating the filtered gradient is as follows: In the formula, Indicates the first The filtering gradient for the nth iteration; Represents the imaging operator; Indicates the first The echo residuals from the nth iteration are transformed into the image domain.

[0032] The formula for calculating the gradient descent result matrix is ​​as follows: In the formula, This represents the gradient descent result matrix; This indicates the iteration step size.

[0033] In S4, refer to Figure 2The flowchart illustrating the thresholding shrinkage method for the gradient descent result matrix, which "thresholds the gradient descent result matrix based on the weight matrix and fractional sparse norm to obtain an updated image matrix," includes the following steps: S41: Extract the gradient descent result matrix row by row and take the absolute value to obtain the magnitude vector of the current row; S42: Extract the weight matrix row by row to obtain the weight vector of the current row; S43: Based on the fractional sparse norm, perform element-wise exponentiation on the product of the weight vector, regularization parameter and iteration step size of the current row to obtain the adaptive threshold vector of the current row. S44: Based on the adaptive threshold vector of the current row, perform threshold shrinking on the magnitude vector of the current row; S45: After threshold shrinking of all rows of the gradient descent result matrix, the updated image matrix is ​​obtained.

[0034] In S41, first traverse each row of the gradient descent result matrix. For the current row being traversed, take the absolute value of each element in the current row and construct the magnitude vector. The extracted expression is: ,in, Indicates row index, This indicates extracting the first [item] from the gradient descent result matrix. Row elements, This indicates taking the absolute value. Indicates the first The magnitude vector of the row.

[0035] In S42, each row of the weight matrix is ​​traversed. For the current row being traversed, the weights of the current row are used to construct a weight vector, and the extracted expression is: ,in, Represents the weight matrix, This indicates extracting the first element from the weight matrix. The weight of the row, Indicates the first The weight vector of the row.

[0036] In S43, when determining the adaptive threshold vector based on the fractional-order sparse norm, the threshold at each position is proportional to the power function of the corresponding weight, iteration step size, and regularization parameter. The power exponent and the proportionality constant are determined by the order of the fractional-order norm. Since the L2 / 3 norm (also known as the 3 / 2 L-norm) achieves a good balance between sparsity and algorithm convergence, being sparser than L1, easier to solve than lower-order norms like L1 / 2, and faster to converge than higher-order norms like L3 / 4, this scheme chooses the L2 / 3 norm for regularization constraints. When the L2 / 3 norm is chosen, the formula for calculating the adaptive threshold vector is as follows: In the formula, Indicates the first The adaptive threshold vector for the row; Represents the regularization parameter; Represents a proportionality constant; It represents the power exponent.

[0037] In S44, refer to Figure 3 The flowchart illustrating the threshold shrinking method for magnitude vectors shows that "threshold shrinking of the magnitude vector of the current row based on the adaptive threshold vector of the current row" includes the following steps: S441: Compare each element in the magnitude vector of the current row with the threshold at the corresponding position in the adaptive threshold vector of the current row; S442: Mark elements greater than the corresponding threshold as valid elements, and set the rest of the elements to zero; S443: Obtain the shrinkage coefficient based on the fractional sparse norm; S444: Amplitude shrinkage of effective elements based on shrinkage coefficient; S445: The effective elements and zero-value elements after amplitude contraction constitute the amplitude vector of the current row after contraction.

[0038] In S441 and S442, each column of the magnitude vector and the adaptive threshold vector of the current row is traversed. The magnitude elements at the same position are compared with the threshold one by one. If the element is greater than the corresponding threshold, the column index is recorded in the valid index set to complete the marking of the valid element; if the element is less than or equal to the corresponding threshold, the element at that position is directly set to zero.

[0039] In S443, the formula for calculating the shrinkage coefficient is as follows: In the formula, Indicates column index; Indicates the first Line number The coefficient of contraction of the column; The matrix representing the gradient descent result matrix contains the first... Line number Valid elements of the column; The matrix representing the gradient descent result matrix contains the first... Line number Intermediate variables for the valid elements of a column; Represents the weight matrix of the first... Line number The weight of the column.

[0040] The formula for calculating the intermediate variables is as follows: In the formula, Represents the hyperbolic cosine function; This represents the inverse hyperbolic cosine function.

[0041] In S444, the formula for amplitude contraction of the effective element is as follows: In the formula, Indicates the first Line number Effective elements after column amplitude contraction.

[0042] In S445, for the magnitude vector of the current row, the magnitude of its valid elements is shrunk and the invalid elements are set to zero, and the shrunk magnitude vector of the current row is completed. At this time, the shrunk magnitude vector of the current row contains both the valid elements after magnitude shrunk and the zero elements, and the order of the elements remains unchanged.

[0043] In S45, after threshold shrinking of all rows of the gradient descent result matrix, the magnitude vectors of each row after shrinkage are integrated to form the updated image matrix. The overall update formula for the image matrix can be adopted. To express, among which The image matrix update function represents the threshold shrinkage operation described above. This method replaces traditional L1 or L1 / 2 regularization with weighted L2 / 3 norm constraints, achieving a sparsity enhancement effect closer to the L0 norm and effectively improving the reconstruction accuracy of single-bit SAR imaging. Through weighted adaptive threshold shrinkage, it combines single-bit quantization and delta modulation to achieve adaptive quantization processing, suppressing noise interference while preserving key target features. The resulting updated image matrix provides a better initial estimate for subsequent iterative reconstruction, thus improving the overall noise resistance and reconstruction quality of the imaging method.

[0044] In S5, refer to Figure 4 The flowchart illustrating the weight matrix update method shows that "updating the weight matrix based on the updated image matrix" includes the following steps: S51: Extract the absolute value of each element from the updated image matrix, add the absolute value of the extracted elements to a preset constant, and take the reciprocal to obtain the updated elements; S52: Arrange all updated elements in their original positions to form the updated weight matrix.

[0045] The formula for calculating the updated element is as follows: In the formula, Indicates the first Line number Update elements in the column; This represents a preset constant, which has a very small value.

[0046] The overall update formula for the weight matrix can be adopted as follows: To express, among which, This represents the weight matrix update function, referring to the update operation described above.

[0047] In S6, refer to Figure 5 The flowchart illustrating the threshold matrix update method shows that "updating the threshold matrix using the δ modulation algorithm" includes the following steps: S61: In the first iteration, the simulated echo signal matrix obtained after transforming the image matrix to the echo domain is separated into a real part signal matrix and an imaginary part signal matrix; S62: Obtain the sign deviation between the real part signal matrix and the imaginary part signal matrix and the current threshold matrix, and update the threshold matrix according to the sign deviation and the initial quantization step size; S63: In non-first iterations, update the quantization step size based on the consistency of the sign deviation between the two iterations; S64: Update the threshold matrix based on the updated quantization step size and the current sign bias.

[0048] In S61-S62, in the first In the number of iterations (of which) The simulated echo signal matrix obtained after transforming the image matrix to the echo domain is: , The real part of the signal matrix is , The imaginary part of the signal matrix is , ,in, Represents the real part of a complex number. The imaginary part of a complex number is represented by two operators that decompose a complex number into two real components.

[0049] Sign bias includes real part sign bias and imaginary part sign bias. The formula for calculating the real part sign bias is as follows: In the formula, Indicates the first The sign deviation of the real part in the number of iterations; Indicates the first The real part of the signal matrix after the nth iteration; Indicates the first The real part of the threshold matrix for the nth iteration.

[0050] The formula for calculating the sign deviation of the imaginary part is as follows: In the formula, Indicates the first Sign deviation of the imaginary part in the number of iterations; Indicates the first The imaginary part of the signal matrix after the nth iteration; Indicates the first The imaginary part of the threshold matrix for the number of iterations.

[0051] In the In the number of iterations (of which) The update formula for the real part of the threshold matrix is ​​as follows: In the formula, This represents the real part of the updated threshold matrix; Indicates the first The quantization step size for each iteration.

[0052] The update formula for the imaginary part of the threshold matrix is ​​as follows: In the formula, This represents the imaginary part of the updated threshold matrix.

[0053] In S63-S64, in the first In the number of iterations (of which) The quantization step size includes the real part quantization step size and the imaginary part quantization step size. In non-first iterations, the update of the threshold matrix depends on both the real part quantization step size and the imaginary part quantization step size. The update methods for the real part quantization step size and the imaginary part quantization step size are as follows: if the current imaginary part sign deviation is the same as the previous real part sign deviation, increase the real part quantization step size; if the current imaginary part sign deviation is different from the previous real part sign deviation, decrease the real part quantization step size; if the current imaginary part sign deviation is the same as the previous imaginary part sign deviation, increase the imaginary part quantization step size; if the current imaginary part sign deviation is different from the previous imaginary part sign deviation, decrease the imaginary part quantization step size.

[0054] The formula for updating the real part quantization step size is as follows: In the formula, This indicates the updated real part quantization step size; and Indicates modulation parameters, , ; This indicates that the current sign deviation of the imaginary part is the same as the previous sign deviation of the real part; This indicates that the current sign deviation of the imaginary part is different from the previous sign deviation of the real part.

[0055] The formula for updating the imaginary part quantization step size is as follows: In the formula, This indicates the updated imaginary part quantization step size; This indicates that the current sign deviation of the imaginary part is the same as the previous sign deviation of the imaginary part; This indicates that the current sign deviation of the imaginary part is different from the previous sign deviation of the imaginary part.

[0056] In the In the number of iterations (of which) After obtaining the updated values ​​of the real part quantization step size and the imaginary part quantization step size, that is... , ,at this time, , The update formula for the real part of the threshold matrix is ​​as follows: The update formula for the imaginary part of the threshold matrix is ​​as follows: The overall quantization step size is updated using the following formula for each iteration: In the formula, the updated real part quantization step size and the updated imaginary part quantization step size are combined to obtain the updated overall quantization step size.

[0057] The first two single-bit quantizations are performed using a random initial threshold and a fixed step size. In subsequent iterations, the quantization step size is adaptively increased or decreased based on the sign consistency of the two consecutive quantization results, and the threshold is dynamically updated. This allows the threshold matrix to gradually approximate the unquantized true measurement value, significantly reducing quantization error. Compared to existing single-bit quantization (fixed zero threshold) and time-varying threshold quantization methods based on Gaussian or uniform distributions, this method can adaptively adjust the threshold according to dynamic signal changes, stabilizing the final threshold near the measured value. This indirectly preserves more amplitude information while retaining only symbol information. Through more accurate threshold quantization, the problem of large errors between the measurement vector and the threshold is effectively improved, mitigating the impact of amplitude information loss on the reconstructed image and maintaining high imaging quality even in low signal-to-noise ratio or weak signal environments. The overall update formula for the threshold matrix can be obtained by using... To express, among which, This represents the δ modulation algorithm described above.

[0058] In S7, the updated threshold and weight matrices are used to iteratively reconstruct the image matrix until the maximum number of iterations is reached, thus obtaining the SAR imaging result. This scheme, while retaining only the sign of the echo signal to reduce ADC hardware costs, introduces a weighted L2 / 3 norm as a sparsity constraint. Compared to L1 or L1 / 2 norms, this is closer to the L0 norm, significantly improving the nonlinear sparse representation capability of the solution. By leveraging the δ modulation concept to achieve adaptive dynamic updating of the threshold, single-bit quantization can approximate the true measurement value, effectively compensating for the loss of amplitude information in symbol quantization.

[0059] This paper provides an example where the radar operates at a 5.3 GHz carrier frequency, corresponding to a wavelength of approximately 5.66 cm, belonging to the C-band. This is suitable for stable imaging applications under medium slant range conditions. A linear frequency modulated (LFM) pulse signal is used as the transmitted waveform, with a pulse width of 2.5 μs and a modulation frequency of 20 THz / s. The platform moves uniformly along the azimuth direction at an equivalent velocity of 150 m / s. The slant range of the scene center is 15 km. A zero-look (frontal side-look) imaging mode is used. The system employs a data sampling scale of 1024 × 1024 in both the range and azimuth directions, with an azimuth sampling rate of 100 Hz. Based on the above data, the imaging result using the traditional Chirp Scaling (matched filtering) algorithm is shown in the image below. Figure 6 The image obtained by imaging using the traditional L1-IST algorithm is shown below. Figure 7 The image obtained by imaging using the traditional L1 / 2-IST algorithm is shown below. Figure 8 The image shows the imaging result obtained using the adaptive single-bit quantization SAR imaging method disclosed in this paper. Figure 9Among them, the imaging results obtained by the traditional matched filtering algorithm have very obvious sidelobes. Compared with the matched filtering algorithm, the L1-IST algorithm has a significant effect on suppressing sidelobes, but a small number of sidelobes still exist in the azimuth dimension. The L1 / 2-IST algorithm has a more obvious effect on suppressing sidelobes. The proposed method has the most obvious effect on suppressing sidelobes and is better than the imaging results of the other two threshold iteration algorithms.

[0060] At a signal-to-noise ratio (SNR) of -10 dB, traditional matched filtering algorithms not only suffer from high sidelobes but also experience significant noise impact on the imaging results, leading to background blur. The L1-IST algorithm effectively suppresses background noise and exhibits lower sidelobes. The L1 / 2-IST algorithm demonstrates better suppression performance in the range dimension compared to the L1-IST algorithm. Our proposed method achieves the best sidelobe suppression in both the azimuth and range dimensions. Through Monte Carlo simulation experiments with an SNR range of -15 dB, our method also demonstrates optimal performance in terms of mean square error, peak SNR, and reconstructed SNR.

[0061] As can be seen from the theory and experiments, this method can significantly improve the reconstruction accuracy and noise robustness of existing single-bit SAR imaging algorithms, and achieve high-resolution imaging while maintaining low hardware complexity.

[0062] Reference Figure 10 This disclosure provides an adaptive single-bit quantization SAR imaging system, comprising: The acquisition module 101 is used to acquire the original echo signal matrix, imaging operator and inverse imaging operator, and initialize the image matrix, threshold matrix and weight matrix; The iterative module 102 is used to iteratively reconstruct the image matrix. In each iteration, based on the threshold matrix, the original echo signal matrix and the inverse imaging operator, the image matrix is ​​transformed to the echo domain and then quantized by a single bit to obtain the echo residual. The filtering module 103 is used to perform matched filtering on the echo residual using the imaging operator to obtain the filtering gradient, and to perform gradient descent on the filtering gradient to obtain the gradient descent result matrix. The shrinkage module 104 is used to perform threshold shrinkage on the gradient descent result matrix based on the weight matrix and the fractional sparse norm to obtain the updated image matrix. The first update module 105 is used to update the weight matrix based on the updated image matrix; The second update module 106 is used to update the threshold matrix using the δ modulation algorithm; The imaging module 107 is used to perform the next iteration based on the updated threshold matrix and weight matrix. When the current iteration count reaches the maximum iteration count, the contracted matrix obtained in the last iteration is used as the SAR imaging result.

[0063] The various variations and specific examples of the adaptive single-bit quantization SAR imaging method provided above are also applicable to the adaptive single-bit quantization SAR imaging system provided in this disclosure. Through the foregoing detailed description of the adaptive single-bit quantization SAR imaging method, those skilled in the art can clearly understand the implementation method of the adaptive single-bit quantization SAR imaging system. For the sake of brevity, they will not be described in detail here.

[0064] A computer device according to embodiments of the present disclosure includes a memory and a processor. The memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and / or cache memory. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0065] The processor may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the computer device to perform desired functions. In one embodiment of this disclosure, the processor is used to run computer-readable instructions stored in the memory, causing the computer device to perform all or part of the steps of the adaptive single-bit quantization SAR imaging method of the foregoing embodiments of this disclosure.

[0066] Those skilled in the art will understand that, in order to solve the technical problem of how to achieve a good user experience, this embodiment may also include well-known structures such as communication buses and interfaces, and these well-known structures should also be included within the protection scope of this disclosure.

[0067] like Figure 11 This is a schematic diagram of a computer device provided for an embodiment of the present disclosure. It illustrates a structural schematic diagram suitable for implementing the computer device in the embodiments of the present disclosure. Figure 11 The computer device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0068] like Figure 11As shown, a computer device may include a processor (such as a central processing unit, graphics processing unit, etc.), which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) or programs loaded from storage devices into random access memory (RAM). The RAM also stores various programs and data required for the operation of the computer device. The processor, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0069] Typically, the following devices can be connected to the I / O interface: input devices, such as sensors or visual information acquisition devices; output devices, such as displays; storage devices, such as magnetic tapes or hard drives; and communication devices. Communication devices allow the computer device to communicate wirelessly or wiredly with other devices (such as edge computing devices) to exchange data. Although Figure 11 A computer apparatus with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0070] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processor, all or part of the steps of the adaptive single-bit quantization SAR imaging method of embodiments of this disclosure are performed.

[0071] For a detailed description of this embodiment, please refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.

[0072] A computer-readable storage medium according to embodiments of the present disclosure stores non-transitory computer-readable instructions. When these non-transitory computer-readable instructions are executed by a processor, all or part of the steps of the adaptive single-bit quantization SAR imaging methods of the foregoing embodiments of the present disclosure are performed.

[0073] The aforementioned computer-readable storage media include, but are not limited to: optical storage media (e.g., CD-ROM and DVD), magneto-optical storage media (e.g., MO), magnetic storage media (e.g., magnetic tape or portable hard drive), media with built-in rewritable non-volatile memory (e.g., memory card), and media with built-in ROM (e.g., ROM cartridge).

[0074] For a detailed description of this embodiment, please refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.

[0075] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0076] In this disclosure, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The block diagrams of devices, apparatuses, devices, and systems involved in this disclosure are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as "comprising," "including," "having," etc., are open-ended terms meaning "including but not limited to," and are used interchangeably with them. The terms "or" and "and" as used herein refer to the terms "and / or," and are used interchangeably with them unless the context clearly indicates otherwise. The term "such as" as used herein refers to the phrase "such as but not limited to," and is used interchangeably with it.

[0077] Additionally, as used herein, the "or" used in a list of items beginning with "at least one" indicates a separate list, such that a list of, for example, "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not imply that the described example is preferred or better than other examples.

[0078] It should also be noted that in the systems and methods of this disclosure, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalent solutions to this disclosure.

[0079] Various changes, substitutions, and modifications can be made to the technology described herein without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, events, means, methods, and actions described above. Currently existing or later-developed processes, machines, manufactures, events, means, methods, or actions that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Therefore, the appended claims include such processes, machines, manufactures, events, means, methods, or actions within their scope.

[0080] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0081] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. An adaptive single-bit quantization SAR imaging method, characterized in that, include: Obtain the original echo signal matrix, imaging operator, and inverse imaging operator, and initialize the image matrix, threshold matrix, and weight matrix; The image matrix is ​​reconstructed iteratively. In each iteration, based on the threshold matrix, the original echo signal matrix, and the inverse imaging operator, the image matrix is ​​transformed to the echo domain and then quantized by a single bit to obtain the echo residual. The imaging operator is used to perform matched filtering on the echo residual to obtain the filtering gradient, and gradient descent is performed on the filtering gradient to obtain the gradient descent result matrix. Based on the weight matrix and the fractional sparse norm, the gradient descent result matrix is ​​thresholded and shrunk to obtain the updated image matrix; The weight matrix is ​​updated based on the updated image matrix; The threshold matrix is ​​updated using the delta modulation algorithm; Based on the updated threshold matrix and weight matrix, the next iteration is performed. When the current iteration count reaches the maximum iteration count, the contracted matrix obtained from the last iteration is used as the SAR imaging result.

2. The adaptive single-bit quantization SAR imaging method according to claim 1, characterized in that, The step of thresholding the gradient descent result matrix based on the weight matrix and the fractional sparse norm to obtain the updated image matrix includes: The gradient descent result matrix is ​​extracted row by row and its absolute value is taken to obtain the magnitude vector of the current row; The weight matrix is ​​extracted row by row to obtain the weight vector of the current row; Based on the fractional sparse norm, the product of the weight vector, regularization parameter and iteration step size of the current row is calculated element-wise to obtain the adaptive threshold vector of the current row. Based on the adaptive threshold vector of the current row, threshold shrinkage is performed on the magnitude vector of the current row; After threshold shrinking of all rows of the gradient descent result matrix, the updated image matrix is ​​obtained.

3. The adaptive single-bit quantization SAR imaging method according to claim 2, characterized in that, The threshold shrinking of the magnitude vector of the current row based on the adaptive threshold vector of the current row includes: Compare each element in the magnitude vector of the current row with the threshold at the corresponding position in the adaptive threshold vector of the current row; Mark elements that are greater than the corresponding threshold as valid elements, and set the rest of the elements to zero; Based on the fractional sparse norm, the shrinkage coefficient is obtained; The effective element is subjected to amplitude shrinkage based on the shrinkage coefficient; The effective elements and zero-value elements after amplitude contraction constitute the amplitude vector of the current row after contraction.

4. The adaptive single-bit quantization SAR imaging method according to claim 1, characterized in that, The step of updating the weight matrix based on the updated image matrix includes: The updated image matrix is ​​extracted element by element and its absolute value is taken. The absolute value of the extracted element is added to a preset constant and the reciprocal is taken to obtain the updated element. Arrange all updated elements in their original positions to form the updated weight matrix.

5. The adaptive single-bit quantization SAR imaging method according to claim 1, characterized in that, The step of updating the threshold matrix using the δ modulation algorithm includes: In the first iteration, the simulated echo signal matrix obtained after transforming the image matrix to the echo domain is separated into a real part signal matrix and an imaginary part signal matrix; Obtain the sign deviation between the real part signal matrix and the imaginary part signal matrix and the current threshold matrix, and update the threshold matrix according to the sign deviation and the initial quantization step size; In non-first iterations, the quantization step size is updated based on the consistency of the sign deviation between the two iterations. The threshold matrix is ​​updated based on the updated quantization step size and the current sign deviation.

6. The adaptive single-bit quantization SAR imaging method according to claim 5, characterized in that, The sign bias includes real part sign bias and imaginary part sign bias, and the quantization step size includes real part quantization step size and imaginary part quantization step size; updating the quantization step size based on the consistency of the sign biases of two consecutive steps includes: If the current imaginary part sign deviation is the same as the previous real part sign deviation, then increase the real part quantization step size; If the current imaginary part sign deviation is different from the previous real part sign deviation, then the real part quantization step size is reduced; If the current imaginary part sign deviation is the same as the previous imaginary part sign deviation, then increase the imaginary part quantization step size; If the current imaginary part sign deviation is different from the previous imaginary part sign deviation, then the imaginary part quantization step size is reduced.

7. The adaptive single-bit quantization SAR imaging method according to claim 1, characterized in that, The fractional sparse norm is the L2 / 3 norm.

8. A computer device, characterized in that, The computer device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the adaptive single-bit quantization SAR imaging method according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the adaptive single-bit quantization SAR imaging method according to any one of claims 1-7.

10. A computer program product comprising computer instructions, characterized in that, When executed by a processor, the computer instructions implement the steps of the adaptive single-bit quantization SAR imaging method according to any one of claims 1-7.

Images