A signal enhancement method and device based on time-slicing compressed sensing three-dimensional imaging radar

Abstract

This invention discloses a signal enhancement method and apparatus for time-slice compressed sensing 3D imaging radar. First, an M×N dimensional random binary matrix is ​​generated as the measurement matrix for compressed sampling. The target area is illuminated using a laser source, and repeated sampling and modulation are performed to obtain a sampling matrix. The j-th column of the sampling matrix is ​​extracted to generate a compressed sampling vector for the corresponding time slice. The covariance between the compressed sampling vector and each column of the measurement matrix is ​​calculated to obtain the compressed sampling covariance vector. The covariance between each column of the measurement matrix and all columns is calculated to obtain the covariance measurement matrix. Based on a compressed sensing 2D reconstruction algorithm, the 2D image of each time slice is reconstructed using the covariance measurement matrix. The reconstructed 2D images of all time slices are stitched together to form a complete 3D image. This method and apparatus can solve the problems of low reconstruction accuracy and poor image quality in existing compressed sensing 3D imaging technologies under noisy environments.

Description

Technical Field

[0001] This invention relates to the field of laser three-dimensional imaging technology, and in particular to a signal enhancement method and apparatus for a time-slice compressed sensing three-dimensional imaging radar. Background Technology

[0002] Non-scanning laser 3D imaging radars are classified into direct ranging and indirect ranging types based on their ranging principles. The core component of direct ranging laser 3D imaging radar is an array detector that can directly measure the time-of-flight of photons. The advantages of laser 3D imaging radars based on APD array detectors lie primarily in their high sensitivity, ability to operate at very low signal levels, and multi-target detection capabilities. However, their development is limited by overall industrial capabilities, resulting in relatively slow research progress and thus, currently, achieving high resolution is not possible.

[0003] In recent years, the emerging single-pixel imaging method based on compressed sensing theory has provided a new approach for low-resolution imaging devices to acquire high-resolution images. This method introduces a spatial light modulator to modulate the spatial distribution of the emitted light or the image of the target. A single-pixel detector collects the total echo energy, continuously changing the modulation template and acquiring corresponding echo pulses. Finally, the target image is obtained through reconstruction calculation. The spatial resolution of the image obtained through this method is no longer limited by the detection device, but depends on the spatial light modulator itself. Furthermore, the single-pixel imaging method based on compressed sensing theory can reconstruct the target image from random measurement signals at a rate far lower than the Nyquist sampling rate requirement, significantly reducing the amount of measurement data required for target reconstruction, thereby further reducing the system's storage pressure and improving acquisition speed. However, since a single-pixel detector can only acquire one echo waveform in a single measurement, it cannot directly obtain the distance information of the entire target. Therefore, traditional single-pixel imaging methods can only obtain a two-dimensional image of the target. Thus, how to extract the target's distance information from echo signals from multiple measurements has become a key problem that urgently needs to be solved in the field of single-pixel three-dimensional imaging. Summary of the Invention

[0004] The purpose of this invention is to provide a signal enhancement method and apparatus for time-slice compressed sensing 3D imaging radar. This method and apparatus can solve the problems of low reconstruction accuracy and poor image quality of existing compressed sensing 3D imaging technology in noisy environments, enabling time-slice compressed sensing 3D imaging radar to stably extract echo information under narrow pulse conditions and noisy environments, enhancing the system's ability to reconstruct the target scene, and achieving high-quality 3D imaging.

[0005] The objective of this invention is achieved through the following technical solution:

[0006] A signal enhancement method based on time-slice compressed sensing three-dimensional imaging radar, the method comprising:

[0007] Step 1: First, generate an M×N dimensional random binary matrix Φ as the measurement matrix for compressed sampling. Use the i-th row of the measurement matrix as the modulation template, where i = 1, 2, ..., M; M is the number of rows in the measurement matrix, which is any value less than N; N is the number of columns in the measurement matrix, which is equal to the number of pixels of the spatial optical modulator used.

[0008] Step 2: Illuminate the target area with a laser light source, and use the echo signal of the reflected light after being modulated by the modulation template as the i-th row of the sampling matrix Y. Repeat the sampling and modulation until the projection of all rows in the measurement matrix is ​​completed to obtain the sampling matrix Y.

[0009] Step 3: Extract the j-th column of the sampling matrix Y to generate the compressed sampling vector y under the corresponding time slice;

[0010] Step 4: Calculate the covariance between the compressed sampling vector y and each column of the measurement matrix to obtain the compressed sampling covariance vector;

[0011] Step 5: Calculate the covariance between each column of the measurement matrix and all columns to obtain the covariance measurement matrix;

[0012] Step 6: For the compressed sampling covariance vector of each time slice in the sampling matrix Y, reconstruct the two-dimensional image of each time slice based on the compressed sensing two-dimensional reconstruction algorithm through the covariance measurement matrix;

[0013] Step 7: Stitch together the reconstructed 2D images from all time slices into a complete 3D image to enhance the signal of the entire 3D image.

[0014] A signal enhancement device based on time-slice compressed sensing three-dimensional imaging radar, the device comprising a laser source, a beam shaping system, a receiving optical system, a photodetector, a spatial optical modulator, a processor, and a memory, wherein:

[0015] The laser source outputs laser pulses with a fixed pulse width, which are collimated, expanded, and homogenized by the beam shaping system before being irradiated onto the target area to be imaged.

[0016] The laser echo reflected from the target area is focused by the receiving optical system and modulated by the spatial optical modulator before illuminating the photodetector. The modulation process of the spatial optical modulator is as follows: an M×N dimensional random binary matrix Φ is generated as the measurement matrix for compressed sampling; the i-th row of the measurement matrix is ​​used as the modulation template, where i = 1, 2, ..., M; the echo signal of the laser echo modulated by the modulation template is the i-th row of the sampling matrix Y; and the projection of all rows in the measurement matrix is ​​repeatedly sampled and modulated to obtain the sampling matrix Y.

[0017] The photodetector sends the acquired echo signal to the processor, which then performs signal enhancement calculations on the echo signal.

[0018] The processor performs signal enhancement calculations on the echo signals as follows: First, all echo signals are extracted, i.e., the j-th column of the sampling matrix Y, to generate a compressed sampling vector y for the corresponding time slice; the covariance between the compressed sampling vector y and each column of the measurement matrix is ​​calculated to obtain a compressed sampling covariance vector; then, the covariance between each column of the measurement matrix and all columns is calculated to obtain a covariance measurement matrix; for the compressed sampling covariance vector of each time slice in the sampling matrix Y, the two-dimensional image of each time slice is reconstructed using the covariance measurement matrix based on the compressed sensing two-dimensional reconstruction algorithm; finally, the reconstructed two-dimensional images of all time slices are stitched together to form a complete three-dimensional image, thereby achieving signal enhancement of the entire three-dimensional image.

[0019] The memory is used to store the echo signal of the photodetector.

[0020] As can be seen from the technical solutions provided by the present invention, the above-mentioned methods and devices can solve the problems of low reconstruction accuracy and poor image quality of existing compressed sensing three-dimensional imaging technology in noisy environments, enabling time-slice compressed sensing three-dimensional imaging radar to stably extract echo information under narrow pulse conditions and noisy environments, enhancing the system's ability to reconstruct the target scene, and achieving high-quality three-dimensional imaging. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the signal enhancement method for a time-slice compressed sensing three-dimensional imaging radar provided in an embodiment of the present invention;

[0023] Figure 2 This is a schematic diagram of the device described in an embodiment of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments, and do not constitute a limitation of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0025] like Figure 1 The diagram shown is a schematic flowchart of a signal enhancement method for a time-slice compressed sensing three-dimensional imaging radar provided in an embodiment of the present invention. The method includes:

[0026] Step 1: First, generate an M×N dimensional random binary matrix Φ as the measurement matrix for compressed sampling. Use the i-th row of the measurement matrix as the modulation template, where i = 1, 2, ..., M; M is the number of rows in the measurement matrix, which is any value less than N, depending on the manually set compression sampling rate; N is the number of columns in the measurement matrix, which is equal to the number of pixels of the spatial optical modulator used.

[0027] In this step, the measurement matrix in this embodiment is a random binary matrix, which is only a specific form of measurement matrix that meets the requirements of compressed sensing theory. Measurement matrices with the same or similar methods, such as random Gaussian matrices and Toeplitz cyclic matrices, are still included within the scope of protection of this invention.

[0028] In practice, the number of rows M is less than the number of columns N, and their ratio M / N is the compression sampling rate in compressed sensing theory. For example, if M = 8192 and N = 65536, the compression sampling rate in this case is 12.5%, and the image resolution is 256*256.

[0029] Step 2: Illuminate the target area with a laser light source, and use the echo signal of the reflected light after being modulated by the modulation template as the i-th row of the sampling matrix Y. Repeat the sampling and modulation until the projection of all rows in the measurement matrix is ​​completed to obtain the sampling matrix Y.

[0030] Step 3: Extract the j-th column of the sampling matrix Y to generate the compressed sampling vector y under the corresponding time slice;

[0031] Step 4: Calculate the covariance between the compressed sampling vector y and each column of the measurement matrix to obtain the compressed sampling covariance vector;

[0032] In this step, the k-th column φ of the measurement matrix is ​​first extracted. k Calculate the compressed sampling vector y and φk The covariance is expressed as:

[0033] Repeat the calculation until the covariance between the compressed sampling vector y and all columns of the measurement matrix has been calculated, to obtain the compressed sampling covariance vector y. cov .

[0034] Step 5: Calculate the covariance between each column of the measurement matrix and all columns to obtain the covariance measurement matrix;

[0035] In this step, the covariance between each column of the measurement matrix and all columns is calculated to obtain the covariance measurement matrix Φ. cov , represented as:

[0036]

[0037] Where, φ j represents the j-th column of the measurement matrix, j = 1, 2, ..., N; cov represents taking the covariance.

[0038] Step 6: For the compressed sampling covariance vector of each time slice in the sampling matrix Y, reconstruct the two-dimensional image of each time slice based on the compressed sensing two-dimensional reconstruction algorithm through the covariance measurement matrix;

[0039] In this step, the compressed sensing two-dimensional reconstruction algorithm is used to reconstruct y. cov =Φ cov Solving for x yields the 2D reconstructed image x under the corresponding time slice; where y cov Φ is the compressed sampling covariance vector calculated in step 4; cov The covariance measurement matrix is ​​obtained from step 5;

[0040] Repeat the above calculations until the two-dimensional image of each time slice within the sampling matrix Y is reconstructed.

[0041] In specific implementations, the compressed sensing two-dimensional reconstruction algorithm includes: augmented Lagrange and alternating direction minimum total variation (TVAL3), block compressed sensing-smooth projection reconstruction (BCS-SPL), two-step iterative shrinkage thresholding (TWIST), or approximate message passing (AMP). Other methods that can achieve the same or similar effects as the two-dimensional image reconstruction method described in this invention are also feasible.

[0042] Step 7: Stitch together the reconstructed 2D images from all time slices into a complete 3D image to enhance the signal of the entire 3D image.

[0043] Based on the above method, embodiments of the present invention also provide a signal enhancement device based on a time-slice compressed sensing three-dimensional imaging radar, such as... Figure 2 The diagram shown is a structural schematic of the device according to an embodiment of the present invention. The device includes a laser source, a beam shaping system, a receiving optical system, a photodetector, a spatial optical modulator, a processor, and a memory, wherein:

[0044] The laser source outputs laser pulses with a fixed pulse width, which are collimated, expanded, and homogenized by the beam shaping system before being irradiated onto the target area to be imaged.

[0045] The laser echo reflected from the target area is focused by the receiving optical system and modulated by the spatial optical modulator before illuminating the photodetector. The modulation process of the spatial optical modulator is as follows: an M×N dimensional random binary matrix Φ is generated as the measurement matrix for compressed sampling; the i-th row of the measurement matrix is ​​used as the modulation template, where i = 1, 2, ..., M; the echo signal of the laser echo modulated by the modulation template is the i-th row of the sampling matrix Y; and the projection of all rows in the measurement matrix is ​​repeatedly sampled and modulated to obtain the sampling matrix Y.

[0046] The photodetector sends the acquired echo signal to the processor, which then performs signal enhancement calculations on the echo signal.

[0047] The processor performs signal enhancement calculations on the echo signals as follows: First, all echo signals are extracted, i.e., the j-th column of the sampling matrix Y, to generate a compressed sampling vector y for the corresponding time slice; the covariance between the compressed sampling vector y and each column of the measurement matrix is ​​calculated to obtain a compressed sampling covariance vector; then, the covariance between each column of the measurement matrix and all columns is calculated to obtain a covariance measurement matrix; for the compressed sampling covariance vector of each time slice in the sampling matrix Y, the two-dimensional image of each time slice is reconstructed using the covariance measurement matrix based on the compressed sensing two-dimensional reconstruction algorithm; finally, the reconstructed two-dimensional images of all time slices are stitched together to form a complete three-dimensional image, thereby achieving signal enhancement of the entire three-dimensional image.

[0048] The memory is used to store the echo signal of the photodetector.

[0049] In specific implementation, the photodetector is a high-sensitivity single-pixel detector with high sensitivity, response speed, and bandwidth, capable of capturing and outputting a relatively complete laser pulse waveform, with a corresponding fixed field of view. Multiple signal enhancement subsystems based on the specific embodiment, implemented using an array detector in a field-of-view manner, are also within the scope of this invention.

[0050] The laser source is a pulsed source, and the laser pulses emitted have a fixed pulse width and emission frequency. The beam shaping system is used to collimate, expand, and homogenize the emitted light from the laser source so that the light intensity irradiated on different target areas is approximately the same.

[0051] In addition, the processor in this embodiment can be implemented by an industrial camera with integrated data acquisition and output functions, an x86 or other architecture processor, an FPGA chip or DSP chip with digital signal acquisition and processing functions, etc.

[0052] It is worth noting that the contents not described in detail in the embodiments of the present invention belong to the prior art known to those skilled in the art.

[0053] In summary, the method and apparatus described in the embodiments of the present invention can enhance the image signals acquired by time-slice compressed sensing 3D imaging radar. They can operate normally and obtain high-quality 3D reconstructed images even under low compression sampling rate conditions and noisy environments. By comparing the results of the time-slice compressed sensing 3D imaging radar at a compression sampling rate of 12.5% ​​with those obtained using the method and apparatus described in the embodiments of the present invention, it can be seen that: in different time slices, the reconstruction results obtained by the method and apparatus described in the embodiments of the present invention are significantly better than those obtained by general methods; in time slices with high noise levels, the method and apparatus described in the embodiments of the present invention have a significant enhancement effect on the echo signal.

[0054] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims. The information disclosed in the background section is intended only to enhance the understanding of the overall background technology of the present invention and should not be construed as an admission or implication in any way that such information constitutes prior art known to those skilled in the art.

Claims

1. A signal enhancement method based on time-slice compressed sensing three-dimensional imaging radar, characterized in that, The method includes: Step 1: First, generate an M×N dimensional random binary matrix Φ as the measurement matrix for compressed sampling. Use the i-th row of the measurement matrix as the modulation template, where i = 1, 2, ..., M; M is the number of rows in the measurement matrix, which is any value less than N; N is the number of columns in the measurement matrix, which is equal to the number of pixels of the spatial optical modulator used. Step 2: Illuminate the target area with a laser light source, and use the echo signal of the reflected light after being modulated by the modulation template as the i-th row of the sampling matrix Y. Repeat the sampling and modulation until the projection of all rows in the measurement matrix is ​​completed to obtain the sampling matrix Y. Step 3: Extract the j-th column of the sampling matrix Y to generate the compressed sampling vector y under the corresponding time slice; Step 4: Calculate the covariance between the compressed sampling vector y and each column of the measurement matrix to obtain the compressed sampling covariance vector; Step 5: Calculate the covariance between each column of the measurement matrix and all columns to obtain the covariance measurement matrix; Step 6: For the compressed sampling covariance vector of each time slice in the sampling matrix Y, reconstruct the two-dimensional image of each time slice based on the compressed sensing two-dimensional reconstruction algorithm through the covariance measurement matrix; Step 7: Stitch together the reconstructed 2D images from all time slices into a complete 3D image to enhance the signal of the entire 3D image.

2. The signal enhancement method for time-slice compressed sensing three-dimensional imaging radar according to claim 1, characterized in that, In step 1, the number of rows M is less than the number of columns N, and their ratio M / N is the compressed sampling rate in compressed sensing theory.

3. The signal enhancement method for time-slice compressed sensing three-dimensional imaging radar according to claim 1, characterized in that, The process of step 4 is as follows: Extract the k-th column φ of the measurement matrix k Calculate the compressed sampling vector y and φ k The covariance is expressed as: Repeat the calculation until the covariance between the compressed sampling vector y and all columns of the measurement matrix has been calculated, to obtain the compressed sampling covariance vector y. cov .

4. The signal enhancement method for time-slice compressed sensing three-dimensional imaging radar according to claim 1, characterized in that, In step 5, the covariance between each column of the measurement matrix and all columns is calculated to obtain the covariance measurement matrix Φ. cov , represented as: Where, φ j represents the j-th column of the measurement matrix, j = 1, 2, ..., N; cov represents taking the covariance.

5. The signal enhancement method for time-slice compressed sensing three-dimensional imaging radar according to claim 1, characterized in that, In step 6, the compressed sensing two-dimensional reconstruction algorithm is used to reconstruct y. cov =Φ cov Solving for x yields the 2D reconstructed image x under the corresponding time slice; where y cov Φ is the compressed sampling covariance vector calculated in step 4; cov The covariance measurement matrix is ​​obtained from step 5; Repeat the above calculations until the two-dimensional image of each time slice within the sampling matrix Y is reconstructed.

6. The signal enhancement method for time-slice compressed sensing three-dimensional imaging radar according to claim 1 or 5, characterized in that, The compressed sensing two-dimensional reconstruction algorithm includes: augmented Lagrange and alternating direction minimum total variation method, block compressed sensing-smooth projection reconstruction method, two-step iterative shrinkage threshold method or approximate message passing method.

7. A signal enhancement device based on time-slice compressed sensing three-dimensional imaging radar, characterized in that, The device includes a laser source, a beam shaping system, a receiving optical system, a photodetector, a spatial optical modulator, a processor, and a memory, wherein: The laser source outputs laser pulses with a fixed pulse width, which are collimated, expanded, and homogenized by the beam shaping system before being projected onto the target area to be imaged. The laser echo reflected from the target area is focused by the receiving optical system and modulated by the spatial optical modulator before illuminating the photodetector. The modulation process of the spatial optical modulator is as follows: an M×N dimensional random binary matrix Φ is generated as the measurement matrix for compressed sampling; the i-th row of the measurement matrix is ​​used as the modulation template, where i = 1, 2, ..., M; the echo signal of the laser echo modulated by the modulation template is the i-th row of the sampling matrix Y; and the projection of all rows in the measurement matrix is ​​repeatedly sampled and modulated to obtain the sampling matrix Y. The photodetector sends the acquired echo signal to the processor, which then performs signal enhancement calculations on the echo signal. The processor performs signal enhancement calculations on the echo signals as follows: First, all echo signals are extracted, i.e., the j-th column of the sampling matrix Y, to generate a compressed sampling vector y for the corresponding time slice; the covariance between the compressed sampling vector y and each column of the measurement matrix is ​​calculated to obtain a compressed sampling covariance vector; then, the covariance between each column of the measurement matrix and all columns is calculated to obtain a covariance measurement matrix; for the compressed sampling covariance vector of each time slice in the sampling matrix Y, the two-dimensional image of each time slice is reconstructed using the covariance measurement matrix based on the compressed sensing two-dimensional reconstruction algorithm; finally, the reconstructed two-dimensional images of all time slices are stitched together to form a complete three-dimensional image, thereby achieving signal enhancement of the entire three-dimensional image. The memory is used to store the echo signal of the photodetector.

8. The signal enhancement device for a time-slice compressed sensing three-dimensional imaging radar as described in claim 7, characterized in that, The laser source is a pulsed source, and the laser pulses emitted have a fixed pulse width and emission frequency. The beam shaping system is used to collimate, expand, and homogenize the emitted light from the laser source so that the light intensity irradiated on different target areas is approximately the same.

Images