A method for anti-motion-blur single-pixel imaging

Abstract

The application discloses an anti-motion-blur single-pixel imaging method, comprising the following steps: S1, initializing the cumulative positioning number as 1, generating a geometric moment mode and a Fourier mode; S2, sequentially projecting the geometric moment mode and the Fourier mode to position a moving object; S3, judging whether the cumulative positioning number reaches a preset maximum positioning number; S4, calculating the position information of the moving object according to the geometric moment value; S5, calculating a compensation displacement according to the position information of the moving object; and S6, reconstructing an anti-motion-blur image of the moving object according to the compensation displacement. The application discloses an anti-motion-blur single-pixel imaging method, which sequentially projects the geometric moment mode and the Fourier mode to position the moving object, greatly reduces the sampling number required for each positioning, and is irrelevant to the positioning accuracy and the sampling number, so that clear imaging of the moving object is realized quickly, and the application can be widely applied to the fields of quantum sensing and quantum imaging industrialization.

Description

Technical Field

[0001] This invention relates to the technical field of quantum sensing and quantum imaging, and specifically to a single-pixel imaging method that resists motion blur. Background Technology

[0002] Single-pixel imaging (SPI) is a derivative of quantum correlation imaging. It simplifies quantum correlation imaging with dual optical paths to a single optical path, uses spatial scanning technology to improve the speed of quantum imaging, and combines compressed sensing algorithms to improve the quality of quantum imaging.

[0003] like Figure 1 For existing single-pixel imaging systems, single-pixel imaging technology can be divided into two types according to modulation methods: one is the pre-modulation-based imaging method, which uses a spatial light modulator to encode light sources such as LEDs and lasers, and then projects the encoded light field onto the scene or object using a lens to encode and modulate the scene information. The modulated light is then collected by a single-pixel detector to obtain one-dimensional light field acquisition information. Finally, a computer uses image reconstruction algorithms such as compressed sensing to reconstruct the scene image from the acquired information. The other is the post-modulation-based imaging method, which uses a lens to image the scene onto a spatial light modulator such as a DMD, then uses the DMD to encode and modulate the scene image, and then combines this with a back-end single-pixel detector to collect the light intensity information of the encoded image. The acquired information is then used by a computer to reconstruct the scene image using image reconstruction algorithms.

[0004] Traditional imaging techniques can shorten single-exposure time, rendering moving objects in an equivalent state of stillness, thus recording clear images. However, single-pixel imaging technology struggles to reduce single-frame imaging time to this level while simultaneously demanding high image quality. Therefore, current single-pixel imaging technology is unable to quickly produce clear images of moving objects. Summary of the Invention

[0005] To address the problem that current single-pixel imaging technology struggles to quickly and clearly image moving objects, this invention provides a single-pixel imaging method that resists motion blur.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A single-pixel imaging method to resist motion blur includes the following steps:

[0008] S1: Initialize the cumulative positioning count to 1, and generate the geometric moment pattern and Fourier pattern;

[0009] S2: Project the geometric moment mode and Fourier mode sequentially to locate the moving object;

[0010] S3: Determine whether the cumulative number of location attempts has reached the preset maximum number of location attempts;

[0011] If so, obtain the geometric moment value and proceed to step S4;

[0012] If not, the cumulative number of location attempts is increased by 1, and the process returns to step S2;

[0013] S4: Calculate the position information of the moving object based on the geometric moment values;

[0014] S5: Calculate the compensation displacement based on the position information of the moving object;

[0015] S6: Reconstruct an anti-motion blur image of the moving object based on the compensated displacement.

[0016] In the above scheme, the moving object is located by projecting geometric moment mode and Fourier mode in sequence, which greatly reduces the number of samplings required for each positioning. The positioning accuracy is independent of the number of samplings, thereby achieving rapid and clear imaging of the moving object.

[0017] Preferably, the geometric moment mode includes the zeroth moment mode. First-order moment x-direction mode and first-order moment y-direction mode ;in,

[0018]

[0019]

[0020]

[0021] in, This represents the number of pixels in the image.

[0022] Preferably, in a single positioning operation, two geometric moment patterns are projected first, followed by continuous projection. A Fourier mode, ≥2.

[0023] Preferably, the two geometric moment modes of the projection include one zeroth-order moment mode. and a first-order moment x-direction mode Or a first-order moment y-direction mode .

[0024] Preferably, 2≤ ≤10.

[0025] Preferably, the positioning frequency is determined by the following formula. :

[0026]

[0027] in, This is the highest projection frequency of the DMD modulator.

[0028] Preferably, the maximum number of positioning attempts is calculated using the following formula. :

[0029]

[0030] by Taking a pixel image as an example, reconstruct a... Number of projections required for a pixel image for:

[0031]

[0032] in, The compression ratio is... .

[0033] Preferably, the position information of the moving object is calculated using the following formula. :

[0034]

[0035]

[0036] in, This refers to the position information in the x-direction. This refers to the position information in the y-direction. These are the geometric moment values ​​obtained by projecting the zeroth-order moment mode. The geometric moment values ​​are obtained by projecting the first-order moment x-direction pattern. The geometric moment value is obtained by projecting the first-order moment y-direction pattern.

[0037] Preferably, the compensation displacement is calculated using the following formula. :

[0038]

[0039]

[0040] in, This represents the compensation displacement in the x-direction. This represents the compensation displacement in the y-direction. Let x be the initial position in the x-direction. Let be the initial position in the y-direction. The number of columns in the image. This represents the number of rows in the image.

[0041] Preferably, the anti-motion blur image of the moving object is reconstructed using the following formula. :

[0042]

[0043] in, For the number of lighting modes, For the first Measurements of the detector under each illumination mode For illuminating the moving object Structured light and These represent coordinates in different directions.

[0044] Beneficial technical effects of the present invention:

[0045] This invention provides a single-pixel imaging method to resist motion blur. By sequentially projecting geometric moment mode and Fourier mode to locate moving objects, the number of samplings required for each positioning is greatly reduced, and the positioning accuracy is independent of the number of samplings, thereby achieving rapid and clear imaging of moving objects. Attached Figure Description

[0046] Figure 1 For existing single-pixel imaging systems;

[0047] Figure 2 This is a flowchart illustrating the implementation steps of the technical solution of the present invention. Detailed Implementation

[0048] 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.

[0049] Example 1

[0050] like Figure 2 As shown, a single-pixel imaging method to resist motion blur includes the following steps:

[0051] S1: Initialize the cumulative positioning count to 1, and generate the geometric moment pattern and Fourier pattern;

[0052] S2: Project the geometric moment mode and Fourier mode sequentially to locate the moving object;

[0053] S3: Determine whether the cumulative number of location attempts has reached the preset maximum number of location attempts;

[0054] If so, obtain the geometric moment value and proceed to step S4;

[0055] If not, the cumulative number of location attempts is increased by 1, and the process returns to step S2;

[0056] S4: Calculate the position information of the moving object based on the geometric moment values;

[0057] S5: Calculate the compensation displacement based on the position information of the moving object;

[0058] S6: Reconstruct an anti-motion blur image of the moving object based on the compensated displacement.

[0059] In practice, the moving object is located by sequentially projecting geometric moment mode and Fourier mode, which greatly reduces the number of samplings required for each positioning. The positioning accuracy is independent of the number of samplings, thus enabling rapid and clear imaging of the moving object.

[0060] Example 2

[0061] A single-pixel imaging method to resist motion blur includes the following steps:

[0062] S1: Initialize the cumulative positioning count to 1, and generate the geometric moment pattern and Fourier pattern;

[0063] In actual implementation, for images ,That The geometric moments of order are:

[0064] ,

[0065] The geometric moment pattern is set using the above formula and generated via computer programming;

[0066] Furthermore, by utilizing the Fourier transform algorithm, a series of Fourier bases with different frequencies and phases are generated as Fourier modes based on the parameters and requirements of the imaging system.

[0067] More specifically, the geometric moment modes include zero-order moment modes. First-order moment x-direction mode and first-order moment y-direction mode ;in,

[0068]

[0069]

[0070]

[0071] in, This represents the number of pixels in the image.

[0072] S2: Project the geometric moment mode and Fourier mode sequentially to locate the moving object;

[0073] More specifically, in a single positioning operation, two geometric moment patterns are first projected, followed by continuous projection. A Fourier mode, ≥2.

[0074] More specifically, the positioning frequency is determined by the following formula. :

[0075]

[0076] in, This is the highest projection frequency of the DMD modulator.

[0077] More specifically, the two geometric moment modes of the projection include one zeroth moment mode. and a first-order moment x-direction mode Or a first-order moment y-direction mode .

[0078] In the specific implementation process, a zero-order moment mode is projected. and one first-order moment x-direction mode Position information in the x-direction can be obtained by projecting a zero-order moment pattern. and one first-order moment y-direction mode It can obtain position information in the x-direction.

[0079] More specifically, 2≤ ≤10.

[0080] In practical implementation, time-division multiplexing technology is used, which can flexibly adjust the speed according to objects with different moving speeds. The coefficients are used to optimize system performance and improve imaging efficiency. For high-speed objects, in order to reduce motion blur, increase the positioning frequency, and quickly acquire position information for motion compensation, the coefficients should be reduced. To reduce the projection time for each group of patterns, the value is used. A value of 2-4 is optimal to ensure that a set of pattern projections is completed in a short time. For slow-moving objects, the value can be appropriately increased without affecting imaging efficiency. A value of 7-10 is optimal.

[0081] S3: Determine whether the cumulative number of location attempts has reached the preset maximum number of location attempts;

[0082] If so, obtain the geometric moment value and proceed to step S4;

[0083] If not, the cumulative number of location attempts is increased by 1, and the process returns to step S2;

[0084] More specifically, the maximum number of positioning attempts is calculated using the following formula. :

[0085]

[0086] by Taking a pixel image as an example, reconstruct a... Number of projections required for a pixel image for:

[0087]

[0088] in, The compression ratio is... .

[0089] S4: Calculate the position information of the moving object based on the geometric moment values;

[0090] More specifically, the position information of a moving object is calculated using the following formula. :

[0091]

[0092]

[0093] in, This refers to the position information in the x-direction. This refers to the position information in the y-direction. These are the geometric moment values ​​obtained by projecting the zeroth-order moment mode. The geometric moment values ​​are obtained by projecting the first-order moment x-direction pattern. The geometric moment value is obtained by projecting the first-order moment y-direction pattern.

[0094] S5: Calculate the compensation displacement based on the position information of the moving object;

[0095] More specifically, the anti-motion blur image of a moving object is reconstructed using the following formula. :

[0096]

[0097] in, The number of illumination modes depends on the required image quality and the speed of the moving object. The more illumination modes there are, the more information is provided, and the sharpness and detail of the image will be improved. However, for high-speed moving objects, it is necessary to appropriately reduce the number of illumination modes in order to reduce imaging time and thus improve motion blur.

[0098] For the first Measurements of the detector under each illumination mode;

[0099] For illuminating the moving object A structured light, based on Fourier modes, is generated and controlled by a DMD modulator;

[0100] and These represent coordinates in different directions.

[0101] S6: Reconstruct an anti-motion blur image of the moving object based on the compensated displacement.

[0102] More specifically, the compensation displacement is calculated using the following formula. :

[0103]

[0104]

[0105] in, This represents the compensation displacement in the x-direction. This represents the compensation displacement in the y-direction. Let x be the initial position in the x-direction. Let be the initial position in the y-direction. The number of columns in the image. This represents the number of rows in the image.

[0106] 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 single-pixel imaging method to resist motion blur, characterized in that, Includes the following steps: S1: Initialize the cumulative positioning count to 1, and generate the geometric moment pattern and Fourier pattern; S2: Project the geometric moment mode and Fourier mode sequentially to locate the moving object; S3: Determine whether the cumulative number of location attempts has reached the preset maximum number of location attempts; If so, obtain the geometric moment value and proceed to step S4; If not, the cumulative number of location attempts increases by 1, and the process returns to step S2; S4: Calculate the position information of the moving object based on the geometric moment values; S5: Calculate the compensation displacement based on the position information of the moving object; S6: Reconstruct an anti-motion blur image of the moving object based on the compensated displacement.

2. The single-pixel imaging method for resisting motion blur according to claim 1, characterized in that, The geometric moment modes include zero-order moment modes. First-order moment x-direction mode and first-order moment y-direction mode ;in, ； ； ； in, This represents the number of pixels in the image.

3. The single-pixel imaging method for resisting motion blur according to claim 2, characterized in that, In a single positioning operation, two geometric moment patterns are first projected, followed by continuous projection. A Fourier mode, ≥2.

4. The single-pixel imaging method for resisting motion blur according to claim 3, characterized in that, The projection has two geometric moment modes, including one zeroth moment mode. and a first-order moment x-direction mode Or a first-order moment y-direction mode .

5. The single-pixel imaging method for resisting motion blur according to claim 3, characterized in that, 2≤ ≤10。 6. The single-pixel imaging method for resisting motion blur according to claim 3, characterized in that, The positioning frequency is determined using the following formula. : ； in, This is the highest projection frequency of the DMD modulator.

7. The single-pixel imaging method for resisting motion blur according to claim 2, characterized in that, The maximum number of positioning attempts can be calculated using the following formula. : ； by Taking a pixel image as an example, reconstruct a... Number of projections required for a pixel image for: ； in, The compression ratio is... .

8. The single-pixel imaging method for resisting motion blur according to claim 1, characterized in that, The position information of a moving object can be calculated using the following formula. : ； ； in, This refers to the position information in the x-direction. This refers to the position information in the y-direction. These are the geometric moment values ​​obtained by projecting the zeroth-order moment mode. The geometric moment values ​​are obtained by projecting the first-order moment x-direction pattern. The geometric moment value is obtained by projecting the first-order moment y-direction pattern.

9. The single-pixel imaging method for resisting motion blur according to claim 8, characterized in that, The compensation displacement is calculated using the following formula. : ； ； in, This represents the compensation displacement in the x-direction. This represents the compensation displacement in the y-direction. Let x be the initial position in the x-direction. Let be the initial position in the y-direction. The number of columns in the image. This represents the number of rows in the image.

10. The single-pixel imaging method for resisting motion blur according to claim 9, characterized in that, Reconstruct anti-motion blur images of moving objects using the following formula. : ； in, For the number of lighting modes, For the first Measurements of the detector under each illumination mode For illuminating the moving object Structured light and These represent coordinates in different directions.

Images