A multi-wavelength computational quantum imaging system and method based on singular value decomposition

By optimizing the speckle map through singular value decomposition, the problem of high computational resources in multi-wavelength quantum imaging technology is solved, improving imaging quality and data acquisition efficiency, making it easier to apply in resource-constrained scenarios.

CN123111464APending Publication Date: 2026-11-17NAT QUANTUM COMM (GUANGDONG) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611153670.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-31
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Multi-wavelength computational quantum imaging technology has high computational resource requirements, making it difficult to apply in resource-constrained scenarios. Furthermore, it produces poor image quality at low sampling rates, exhibits significant noise under undersampling conditions, and struggles to retain detailed information when the number of measurements is insufficient.

Method used

Singular value decomposition is used to optimize speckle patterns. By using a multi-wavelength light source, a diffraction grating, a spatial light modulator, and a detection module, combined with the singular value decomposition algorithm, the optimized speckle pattern is generated, reducing computational resource requirements and improving imaging quality.

Benefits of technology

It reduces the sampling time for reconstructed images, improves data acquisition efficiency, reduces computational resource requirements, and significantly improves image quality while keeping the number of measurements constant.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123111464A_ABST
    Figure CN123111464A_ABST
Patent Text Reader

Abstract

This invention discloses a multi-wavelength computational quantum imaging system based on singular value decomposition (SVD), comprising a multi-wavelength light source, a first diffraction grating, a second diffraction grating, a spatial light modulator, an object to be imaged, a detection module, and a computer. This invention also discloses a multi-wavelength computational quantum imaging method based on SVD, implemented using the aforementioned multi-wavelength computational quantum imaging system based on SVD. This invention discloses a multi-wavelength computational quantum imaging system and method based on SVD. By optimizing the speckle pattern using SVD, it reduces the sampling time for reconstructed images, improves data acquisition efficiency, reduces the demand for computational resources, and significantly improves image quality while maintaining the same number of measurements, making multi-wavelength computational quantum imaging easier to promote and apply.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the fields of quantum computing and quantum imaging technology, specifically to a multi-wavelength computational quantum imaging system and method based on singular value decomposition. Background Technology

[0002] Quantum imaging (commonly known as magic quantum imaging or quantum ghost imaging) is one of the key technologies in the field of quantum information technology, possessing significant advantages in high precision and high definition. Computational quantum imaging reconstructs images by utilizing the intensity correlation between an object beam and a reference beam. Its principle is based on a second-order correlation function, employing a bucket detector with no spatial resolution to collect the total intensity information of the object beam and combining it with the spatial modulation mode of the reference beam to reconstruct the spatial distribution of the target object. Compared to traditional optical imaging, computational quantum imaging exhibits unique advantages in low-light conditions, anti-interference capabilities, and information encryption.

[0003] Computational quantum imaging simplifies optical systems by replacing physical reference arms with numerical simulations, further promoting the practical application of this technology. In recent years, multi-wavelength computational quantum imaging (MWCGI) has become a research hotspot. By fusing information from multiple wavelength bands, it can not only recover the geometric structure of a target but also acquire color and spectral features, which has important application value in fields such as medical imaging, remote sensing, and industrial inspection.

[0004] However, the data volume of multi-wavelength computational quantum imaging systems increases exponentially with the number of wavelengths, leading to a sharp increase in computational complexity and storage requirements. Expanding the RGB three-channel to multiple bands further exacerbates the consumption of memory and computing resources. Secondly, multi-wavelength computational quantum imaging technology suffers from poor image quality at low sampling rates, especially under undersampling conditions where noise is significant, and it struggles to retain detailed information when the number of measurements is insufficient. Furthermore, while current optimizations for multi-wavelength computational quantum imaging can improve imaging efficiency, they rely on high-dimensional matrix operations, further increasing the computational burden.

[0005] Therefore, existing multi-wavelength computational quantum imaging technologies have high computational resource requirements, making them difficult to apply in resource-constrained scenarios. Summary of the Invention

[0006] To address the issue of high computational resource requirements in existing multi-wavelength computational quantum imaging technologies, this invention provides a multi-wavelength computational quantum imaging system and method based on singular value decomposition.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] A multi-wavelength computational quantum imaging system based on singular value decomposition includes a multi-wavelength light source, a first diffraction grating, a second diffraction grating, a spatial light modulator, an object to be imaged, a detection module, and a computer.

[0009] The multi-wavelength light source is used to output polychromatic light beams;

[0010] The first diffraction grating is used to split and reflect the polychromatic beam;

[0011] The second diffraction grating is used to reflect and collimate the polychromatic beam after it has been split and reflected.

[0012] The spatial light modulator is used to modulate the aligned beam using the optimized speckle pattern.

[0013] The object to be imaged is positioned between the spatial light modulator and the detection module;

[0014] The detection module is used to detect optical signals of a specified wavelength;

[0015] The computer is used to generate a random speckle pattern and perform singular value decomposition on the random speckle pattern to obtain an optimized speckle pattern; it is also used to obtain a reconstructed image of each band based on the detection results and obtain an image of the object to be imaged through fusion.

[0016] In the above scheme, by optimizing the speckle map using singular value decomposition, the sampling time for reconstructed images is reduced, the efficiency of data acquisition is improved, the demand for computing resources is reduced, and the imaging quality is significantly improved while keeping the number of measurements the same, making multi-wavelength computational quantum imaging easier to promote and apply.

[0017] Preferably, the multi-wavelength light source includes a white light-emitting diode, a beam expander, a pinhole, a collimating lens, and an aperture stop arranged coaxially in sequence.

[0018] Preferably, the first diffraction grating and the second diffraction grating are arranged in parallel, and the second diffraction grating is placed in the first-order diffraction optical path of the first diffraction grating.

[0019] Preferably, the detection module includes a lens, a filter module, and a barrel detector;

[0020] The lens is used to focus the light beam passing through the object to be imaged onto the filter module;

[0021] The filtering module includes three narrowband filters of different wavelengths, each of which is used to transmit light beams of the corresponding wavelength.

[0022] The barrel detector includes three wavelength-selective photodetectors, each of which is used to detect the beam of light in a corresponding wavelength band.

[0023] Preferably, the computer obtains the optimized speckle pattern through the following process:

[0024] The first The random speckle pattern corresponding to the measurement. Expanded row-wise into a one-dimensional row vector, it is represented as:

[0025]

[0026] Among them, random speckle pattern The size is The length of the one-dimensional row vector is x-axis y-axis , ,variable , Three specified wavelengths, To measure the total number of times;

[0027] Will In this measurement, the one-dimensional row vectors are stacked sequentially row by row to form a shape of size [missing information]. random measurement matrix , represented as:

[0028]

[0029] For random measurement matrix Perform singular value decomposition to obtain the optimized measurement matrix, represented as:

[0030]

[0031] in, and Given two orthogonal matrices, For matrix transpose, It is a diagonal matrix;

[0032] choose As the optimized speckle pattern.

[0033] Preferred, , , .

[0034] Preferably, the reconstructed image for each band is represented as follows:

[0035]

[0036] in, The object to be imaged;

[0037] The image of the object to be imaged is obtained by fusing the reconstructed images of each band, and is represented as:

[0038] .

[0039] A multi-wavelength computational quantum imaging method based on singular value decomposition, implemented using the aforementioned multi-wavelength computational quantum imaging system based on singular value decomposition, includes the following steps:

[0040] S1: Outputs polychromatic light beams from multi-wavelength light sources;

[0041] S2: The polychromatic beam is first split and reflected by the first diffraction grating, and then reflected and collimated by the second diffraction grating;

[0042] S3: The spatial light modulator modulates the aligned beam using the optimized speckle pattern; wherein, the optimized speckle pattern is obtained by performing singular value decomposition on the random speckle pattern.

[0043] S4: The modulated beam passes through the object to be imaged and is then input into the detection module;

[0044] S5: The detection module detects the optical signal of a specified wavelength;

[0045] S6: Obtain the reconstructed image of each band based on the detection results, and obtain the image of the object to be imaged through fusion.

[0046] Preferably, in step S3, the optimized speckle pattern is obtained through the following steps:

[0047] S3.1: The first The random speckle pattern corresponding to the measurement. Expanded row-wise into a one-dimensional row vector, it is represented as:

[0048]

[0049] Among them, random speckle pattern The size is The length of the one-dimensional row vector is x-axis y-axis , ,variable , Three specified wavelengths, To measure the total number of times;

[0050] S3.2: Will In this measurement, the one-dimensional row vectors are stacked sequentially row by row to form a shape of size [missing information]. random measurement matrix , represented as:

[0051]

[0052] S3.3: For the random measurement matrix Perform singular value decomposition to obtain the optimized measurement matrix, represented as:

[0053]

[0054] in, and Given two orthogonal matrices, For matrix transpose, It is a diagonal matrix;

[0055] S3.4: Selection As the optimized speckle pattern.

[0056] Preferably, in step S6, the reconstructed image of each band is represented as follows:

[0057]

[0058] in, The object to be imaged;

[0059] The image of the object to be imaged is obtained by fusing the reconstructed images of each band, and is represented as:

[0060] .

[0061] Beneficial technical effects of the present invention:

[0062] This invention provides a multi-wavelength computational quantum imaging system and method based on singular value decomposition. By optimizing the speckle pattern using singular value decomposition, the sampling time for reconstructed images is reduced, the efficiency of data acquisition is improved, and the demand for computing resources is reduced. Moreover, while keeping the number of measurements the same, the imaging quality can be significantly improved, making multi-wavelength computational quantum imaging easier to promote and apply. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the module connection of the present invention;

[0064] Figure 2 This is a schematic diagram of the overall structure of the present invention;

[0065] The components include: 1. Multi-wavelength light source; 11. White light-emitting diode; 12. Beam expander; 13. Pinhole; 14. Collimating lens; 15. Aperture stop; 2. First diffraction grating; 3. Second diffraction grating; 4. Spatial light modulator; 5. Object to be imaged; 6. Detection module; 61. Lens; 62. Filter module; 63. Bucket detector; 7. Computer. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments. However, the scope of protection of this invention is not limited to the specific embodiments described below.

[0067] Example 1

[0068] like Figure 1 As shown, a multi-wavelength computational quantum imaging system based on singular value decomposition includes a multi-wavelength light source 1, a first diffraction grating 2, a second diffraction grating 3, a spatial light modulator 4, an object to be imaged 5, a detection module 6, and a computer 7.

[0069] The multi-wavelength light source 1 is used to output polychromatic light beams;

[0070] The first diffraction grating 2 is used to perform spectral reflection of the polychromatic beam;

[0071] The second diffraction grating 3 is used to reflect and collimate the polychromatic beam after it has been split and reflected.

[0072] The spatial light modulator 4 is used to modulate the aligned beam using the optimized speckle pattern.

[0073] The object to be imaged 5 is positioned between the spatial light modulator 4 and the detection module 6;

[0074] The detection module 6 is used to detect optical signals of a specified wavelength;

[0075] The computer 7 is used to generate a random speckle pattern and perform singular value decomposition on the random speckle pattern to obtain an optimized speckle pattern; it is also used to obtain a reconstructed image of each band based on the detection results and obtain an image of the object 5 to be imaged through fusion.

[0076] In the specific implementation process, by using singular value decomposition to optimize the speckle map, the sampling time for reconstructed images is reduced, the efficiency of data acquisition is improved, the demand for computing resources is reduced, and the imaging quality is significantly improved while keeping the number of measurements the same, making multi-wavelength computational quantum imaging easier to promote and apply.

[0077] Example 2

[0078] like Figure 2 As shown, a multi-wavelength computational quantum imaging system based on singular value decomposition includes a multi-wavelength light source 1, a first diffraction grating 2, a second diffraction grating 3, a spatial light modulator 4, an object to be imaged 5, a detection module 6, and a computer 7.

[0079] More specifically, the multi-wavelength light source 1 includes a white light-emitting diode 11, a beam expander 12, a pinhole 13, a collimating lens 14, and an aperture stop 15 arranged coaxially in sequence.

[0080] In the specific implementation process, the white light emitting diode 11 generates a polychromatic beam. The polychromatic beam is collimated and expanded after passing through the beam expander 12 and the collimator 14. A pinhole 13 is set at the focal position between the beam expander 12 and the collimator 14 to filter the beam. The collimated and expanded polychromatic beam passes through the aperture stop 15 to select the diameter of the incident beam.

[0081] More specifically, the first diffraction grating 2 and the second diffraction grating 3 are arranged in parallel (with the same orientation of the grating lines), and the second diffraction grating 3 is placed in the first-order diffraction optical path of the first diffraction grating 2.

[0082] In the specific implementation process, the angle between the first diffraction grating 2 and the optical axis of the white light emitting diode 11 is 0~30°.

[0083] In the specific implementation process, the polychromatic light beam is incident on the first diffraction grating 2, and after being split and reflected by the first diffraction grating 2, the reflected light is then reflected and collimated by the second diffraction grating 3, so that the light beam is emitted in a parallel form.

[0084] The computer 7 generates a random speckle map and performs singular value decomposition on the random speckle map to obtain an optimized speckle map.

[0085] More specifically, the computer 7 obtains the optimized speckle map through the following process:

[0086] The first The random speckle pattern corresponding to the measurement. Expanded row-wise into a one-dimensional row vector, it is represented as:

[0087]

[0088] Among them, random speckle pattern The size is The length of the one-dimensional row vector is x-axis y-axis , ,variable , Three specified wavelengths, To measure the total number of times;

[0089] Will In this measurement, the one-dimensional row vectors are stacked sequentially row by row to form a shape of size [missing information]. random measurement matrix , represented as:

[0090]

[0091] Random measurement matrix Each row in the matrix corresponds to the light field distribution of a single measurement, and the matrix as a whole characterizes the interaction between the object and the speckle pattern under multiple measurements.

[0092] For random measurement matrix Perform singular value decomposition to obtain the optimized measurement matrix, represented as:

[0093]

[0094] in, and Given two orthogonal matrices, For matrix transpose, It is a diagonal matrix;

[0095] choose As the optimized speckle pattern.

[0096] The computer 7 inputs the optimized speckle pattern into the spatial light modulator 4, which uses the optimized speckle pattern to modulate the light beam. The modulated light beam passes through the object to be imaged 5, carrying the spatial structure information and color information of the object to be imaged 5.

[0097] More specifically, the detection module 6 includes a lens 61, a filter module 62, and a barrel detector 63;

[0098] The lens 61 is used to focus the light beam passing through the object 5 to be imaged onto the filter module 62;

[0099] The filter module 62 includes three different wavelength bands ( , , Narrowband filters, with each band of narrowband filter used to transmit light beams of the corresponding band;

[0100] The barrel detector 63 includes three wavelength-selective photodetectors, each of which is used to detect a beam of light in a corresponding wavelength band.

[0101] More specifically, , , .

[0102] More specifically, the reconstructed image for each band is represented as follows:

[0103]

[0104] in, Object 5 to be imaged;

[0105] The image of the object to be imaged, 5, is obtained by fusing the reconstructed images of each band, and is represented as follows:

[0106] .

[0107] A multi-wavelength computational quantum imaging method based on singular value decomposition, implemented using the aforementioned multi-wavelength computational quantum imaging system based on singular value decomposition, includes the following steps:

[0108] S1: A polychromatic light beam is output from a multi-wavelength light source 1;

[0109] S2: The polychromatic beam is first split and reflected by the first diffraction grating 2, and then reflected and collimated by the second diffraction grating 3;

[0110] S3: The spatial light modulator 4 modulates the aligned beam using the optimized speckle pattern; wherein, the optimized speckle pattern is obtained by performing singular value decomposition on the random speckle pattern.

[0111] S4: The modulated beam passes through the object to be imaged 5 and is then input into the detection module 6;

[0112] S5: The detection module 6 detects the optical signal of the specified wavelength;

[0113] S6: Obtain the reconstructed image of each band based on the detection results, and obtain the image of the object 5 to be imaged by fusion.

[0114] More specifically, in step S3, the optimized speckle pattern is obtained through the following steps:

[0115] S3.1: The first The random speckle pattern corresponding to the measurement. Expanded row-wise into a one-dimensional row vector, it is represented as:

[0116]

[0117] Among them, random speckle pattern The size is The length of the one-dimensional row vector is x-axis y-axis , ,variable , Three specified wavelengths, To measure the total number of times;

[0118] S3.2: Will In this measurement, the one-dimensional row vectors are stacked sequentially row by row to form a shape of size [missing information]. random measurement matrix , represented as:

[0119]

[0120] Random measurement matrix Each row in the matrix corresponds to the light field distribution of a single measurement, and the matrix as a whole characterizes the interaction between the object and the speckle pattern under multiple measurements.

[0121] S3.3: For the random measurement matrix Perform singular value decomposition to obtain the optimized measurement matrix, represented as:

[0122]

[0123] in, and Given two orthogonal matrices, For matrix transpose, It is a diagonal matrix;

[0124] S3.4: Selection As the optimized speckle pattern.

[0125] More specifically, in step S6, the reconstructed image for each band is represented as follows:

[0126]

[0127] in, Object 5 to be imaged;

[0128] The image of the object to be imaged, 5, is obtained by fusing the reconstructed images of each band, and is represented as follows:

[0129] .

[0130] Based on the disclosure and teachings of the foregoing specification, those skilled in the art can make changes and modifications to the above embodiments. Therefore, the present invention is not limited to the specific embodiments disclosed and described above, and some modifications and changes to the invention should also fall within the protection scope of the claims of the present invention. Furthermore, although some specific terms are used in this specification, these terms are only for convenience of explanation and do not constitute any limitation on the present invention.

Claims

1. A multi-wavelength computational quantum imaging system based on singular value decomposition, characterized in that, It includes a multi-wavelength light source, a first diffraction grating, a second diffraction grating, a spatial light modulator, the object to be imaged, a detection module, and a computer; The multi-wavelength light source is used to output polychromatic light beams; The first diffraction grating is used to split and reflect the polychromatic beam; The second diffraction grating is used to reflect and collimate the polychromatic beam after it has been split and reflected. The spatial light modulator is used to modulate the aligned beam using the optimized speckle pattern. The object to be imaged is positioned between the spatial light modulator and the detection module; The detection module is used to detect optical signals of a specified wavelength; The computer is used to generate a random speckle pattern and perform singular value decomposition on the random speckle pattern to obtain an optimized speckle pattern; it is also used to obtain a reconstructed image of each band based on the detection results and obtain an image of the object to be imaged through fusion.

2. The multi-wavelength computational quantum imaging system based on singular value decomposition of claim 1, wherein, The multi-wavelength light source includes a white light-emitting diode, a beam expander, a pinhole, a collimating lens, and an aperture stop arranged coaxially in sequence.

3. The multi-wavelength computational quantum imaging system based on singular value decomposition of claim 1, wherein, The first diffraction grating and the second diffraction grating are arranged in parallel, and the second diffraction grating is placed in the first-order diffraction optical path of the first diffraction grating.

4. The multi-wavelength computational quantum imaging system based on singular value decomposition of claim 1, wherein, The detection module includes a lens, a filter module, and a barrel detector; The lens is used to focus the light beam passing through the object to be imaged onto the filter module; The filtering module includes three narrowband filters of different wavelengths, each of which is used to transmit light beams of the corresponding wavelength. The barrel detector includes three wavelength-selective photodetectors, each of which is used to detect the beam of light in a corresponding wavelength band.

5. The multi-wavelength computational quantum imaging system based on singular value decomposition of claim 1, wherein, The computer obtains the optimized speckle pattern through the following process: The first The random speckle pattern corresponding to the measurement. Expanded row-wise into a one-dimensional row vector, it is represented as: ; Among them, random speckle pattern The size is The length of the one-dimensional row vector is x-axis y-axis , ,variable , Three specified wavelengths, To measure the total number of times; The individual one-dimensional row vectors in the secondary measurement are stacked one after the other in rows to form a random measurement matrix of size , denoted as: ; Random measurement matrix Performing singular value decomposition, the optimized measurement matrix is obtained, denoted as: ; in, and Given two orthogonal matrices, For matrix transpose, It is a diagonal matrix; selecting as the optimized speckle pattern.

6. A multi-wavelength computational quantum imaging system based on singular value decomposition according to claim 5, characterized in that, , , 。 7. The multi-wavelength computational quantum imaging system based on singular value decomposition of claim 5, wherein, The reconstructed image for each band is represented as follows: ; wherein is an object to be imaged; The image of the object to be imaged is obtained by fusing the reconstructed images of each band, and is represented as: 。 8. A multi-wavelength computational quantum imaging method based on singular value decomposition, characterized in that, Includes the following steps: S1: Outputs polychromatic light beams from multi-wavelength light sources; S2: The polychromatic beam is first split and reflected by the first diffraction grating, and then reflected and collimated by the second diffraction grating; S3: The spatial light modulator modulates the aligned beam using the optimized speckle pattern; wherein, the optimized speckle pattern is obtained by performing singular value decomposition on the random speckle pattern. S4: The modulated beam passes through the object to be imaged and is then input into the detection module; S5: The detection module detects the optical signal of a specified wavelength; S6: Obtain the reconstructed image of each band based on the detection results, and obtain the image of the object to be imaged through fusion.

9. The method of claim 8, wherein, In step S3, the optimized speckle pattern is obtained through the following steps: S3.1: The first The random speckle pattern corresponding to the measurement. Expanded row-wise into a one-dimensional row vector, it is represented as: ; Among them, random speckle pattern The size is The length of the one-dimensional row vector is x-axis y-axis , ,variable , Three specified wavelengths, To measure the total number of times; S3.2: Will In this measurement, the one-dimensional row vectors are stacked sequentially row by row to form a shape of size [missing information]. random measurement matrix , is represented as: ; S3.3: For the random measurement matrix Perform singular value decomposition to obtain the optimized measurement matrix, represented as: ; in, and Given two orthogonal matrices, For matrix transpose, It is a diagonal matrix; S3.4: Selection As the optimized speckle pattern.

10. A multi-wavelength computational quantum imaging method based on singular value decomposition according to claim 9, characterized in that, In step S6, the reconstructed image for each band is represented as follows: ; in, The object to be imaged; The image of the object to be imaged is obtained by fusing the reconstructed images of each band, and is represented as: 。