A scattering imaging method based on phase recursion guided by normalized weighting terms

Abstract

This invention discloses a scattering imaging method based on phase recursion guided by normalized weighting terms, comprising the following steps: S1, obtaining the speckle pattern of the unknown target hidden behind the scattering medium using an imaging system capable of high-quality reconstruction of an unknown target image through the scattering medium; S2, segmenting and projecting the speckle pattern and performing triple correlation operations; S3, reconstructing the Fourier phase using a phase recursion algorithm guided by normalized weighting terms; S4, repeating S3, combining the recursively reconstructed Fourier phase with the Fourier amplitude extracted from the speckle correlation, and outputting the reconstructed target image. This method can achieve accurate imaging through the scattering medium.

Description

Technical Field

[0001] This invention relates to the field of optical imaging technology, specifically to a scattering imaging method based on phase recursion guided by normalized weighting terms. Background Technology

[0002] In nature, information-carrying light is scattered when it passes through biological tissue, smog, clouds, or other diffusers. The information it originally carried is then hidden in complex, disordered speckle patterns. Scattering is ubiquitous and causes various problems in daily life. For example, smog severely reduces visibility, easily causing traffic safety issues; and biological tissues hinder imaging of the internal parts of organisms during biomedical diagnosis and treatment. Traditional optical imaging systems cannot obtain target images under scattering conditions. Therefore, how to achieve imaging of unknown targets through scattering media has significant research value in fields such as biomedicine, geological exploration, remote sensing mapping, and military security.

[0003] Based on the principle of speckle autocorrelation, the amplitude of the target image after hiding the scattering medium can be obtained from the autocorrelation of the speckle pattern. Therefore, recovering the phase information of the scattered target is the key to realizing scattering imaging. In phase recovery techniques for speckle patterns, the GS algorithm and the Fienup algorithm both suffer from the problem of losing target orientation information and having weak noise resistance, which restricts their practical application in complex environments. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention proposes a scattering imaging method based on phase recursion guided by normalized weighting terms, which can achieve accurate imaging through scattering media.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] A scattering imaging method based on phase recursion guided by normalized weighting terms includes the following steps:

[0007] S1. Obtain the speckle pattern of the unknown target hidden behind the scattering medium by using an imaging system that reconstructs a high-quality image of the unknown target through the scattering medium.

[0008] S2. Segment and project the speckle pattern, and perform triple correlation operations;

[0009] S3. Reconstruct the Fourier phase using a phase recursive algorithm guided by normalized weighting terms;

[0010] S4. Repeat S3 to combine the recursively reconstructed Fourier phase with the Fourier amplitude extracted by speckle correlation to output the reconstructed target image.

[0011] Preferably, the specific content of S1 includes: the laser light emitted by the laser source is collimated and modulated by rotating frosted glass to obtain pseudo-thermal light, which is then irradiated onto the spatial light modulator. The light carrying the target information passes through the scattering medium and is filtered by a narrow-band filter to remove stray light. It is then collected and scattered by a high-resolution camera to form a speckle pattern.

[0012] The speckle pattern I(x) can be represented as:

[0013] I(x) = O(x) * S(x)

[0014] Where I(x) represents the acquired speckle pattern, O(x) represents the intensity of the target image, S(x) represents the point spread function, and * represents the convolution operation.

[0015] Preferably, the specific content of S2 includes:

[0016] S21. Set the segmentation interval and segmentation overlap rate, and segment the speckle pattern according to this standard, dividing the speckle image into n sub-speckle patterns. Adjacent sub-speckle patterns overlap to ensure they contain all the information of the image. The intensity I of the m-th sub-speckle pattern... m It is given by the following formula:

[0017] I m =O*S m

[0018] Among them, S m It is the point spread function at the center of the m-th sub-speckle.

[0019] S22. All the sub-speckle patterns obtained by segmentation are uniformly projected onto the 0 to π area using Radon transform to obtain one-dimensional sub-speckle projection data.

[0020] S23. Perform triple correlation on the one-dimensional sub-speckle data obtained from the projection. The measure of triple correlation is the bispectrum. The bispectrum obtained by triple correlation of the m-th sub-speckle projection can be expressed as:

[0021]

[0022] Where f1 and f2 are spatial frequencies, It is the position vector in the image. This indicates the conjugate operation.

[0023] S24. Extract the TC phase by calculating the bispectrum of each sub-speckle projection.

[0024] Preferably, the specific content of S3 includes:

[0025] S31. Calculate the weighted average term W in the recursive process. m (i,p) can be represented as:

[0026]

[0027] in, It is the variance of the bispectral modulus, and the weighted average term W. m (i,p) can be viewed as the square of the bispectral signal-to-noise ratio.

[0028] S32. Introduce phasor difference D(f,Δf) and phase difference. As a constraint parameter in the recursive process, the phasor difference can be expressed as:

[0029]

[0030] and phase difference between It can be represented as:

[0031]

[0032] The frequency difference Δf = p can provide information about the phase. Two estimates, one is The other one is

[0033] S33. Introducing a weighted phasor Ω(i) to guide the normalized weighted terms in each iteration of the TC phase recursive Fourier phase algorithm, the weighted phasor at f = i in the q-th iteration of the m-th projection can be expressed as:

[0034]

[0035] Each iteration in the recursive process is performed by normalizing Ω(i). The initial value of the iteration can be the phase estimated by the recursive algorithm or it can be zero.

[0036] S34. The weighted phasor Ω(i) is renormalized in each iteration, and the normalized weighted term R is obtained from the normalization operation. m (i) The condition for determining the iteration loop can be expressed as:

[0037]

[0038] When R m (i) The iteration process terminates when the change in continuous iterations is comparable to the noise.

[0039] The relationship between the S35 and TC phases and the target phase can be expressed as:

[0040]

[0041] The fourfold phase of the speckle TC phase β(f1,f2) is compared with the target Fourier phase. The relationship between them is recursively used in iterative calculations to obtain the target Fourier phase information.

[0042] Preferably, the specific content of S4 includes:

[0043] S41. Repeat S3 to recursively reconstruct the target Fourier phase from the TC phase of all sub-speckle data.

[0044] S42. Perform autocorrelation on the speckle pattern I(x). The calculation process can be expressed as follows:

[0045]

[0046] in The autocorrelation operation is represented by the point spread function, which can be approximated as an impulse function within a certain range. Therefore, the autocorrelation of the speckle pattern can be approximately equivalent to the autocorrelation of the target image, and is expressed as:

[0047]

[0048] S43. According to Wiener-Khinchin's theorem, the Fourier amplitude of the target is calculated using the Fourier transform of the speckle map autocorrelation:

[0049]

[0050] S44. Combine the Fourier phase obtained by recursion with the Fourier amplitude obtained by autocorrelation to output the reconstructed target image.

[0051] This invention has the following characteristics and beneficial effects:

[0052] This invention simplifies the process of triple correlation calculation, effectively reduces redundant operations in the calculation process, and greatly improves the algorithm execution speed.

[0053] This invention introduces triple correlation operations for phase reconstruction, which effectively preserves the orientation information of the original target image and enables high-precision reconstruction of unknown targets.

[0054] This invention introduces a normalized weighted term-guided recursion method, which effectively enhances the algorithm's noise resistance and improves the convergence speed during the iteration process, effectively reducing the number of phase recovery iterations, thereby obtaining better imaging results. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0056] Figure 1 This is a flowchart of a scattering imaging method based on phase recursion guided by normalized weighting terms, provided in an embodiment of the present invention.

[0057] Figure 2 This is a schematic diagram of the structure of an imaging system for high-quality reconstruction of an unknown target image through a scattering medium, provided in an embodiment of the present invention;

[0058] Figure 3 This is the output target image of the spatial light modulator in an imaging system for high-quality reconstruction of an unknown target image through a scattering medium, as provided in this embodiment of the invention.

[0059] Figure 4 This is a speckle pattern acquired by an imaging system provided by the present invention for high-quality reconstruction of an unknown target image through a scattering medium in an embodiment of the present invention;

[0060] Figure 5 The image is a reconstructed image obtained by a scattering imaging method based on phase recursion guided by normalization weighting terms in an embodiment of the present invention.

[0061] In the diagram, 1-laser source, 2-beam expander, 3-rotating frosted glass, 4-variable aperture, 5-spatial light modulator, 6-scattering medium, 7-variable aperture, 8-narrowband filter, 9-camera. Detailed Implementation

[0062] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0063] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0064] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0065] This invention provides a scattering imaging method based on phase recursion guided by normalized weighting terms, such as... Figure 1 As shown, this imaging method includes:

[0066] S1. An imaging system that reconstructs high-quality images of unknown targets through a scattering medium is used to obtain speckle patterns formed by scattering of unknown targets hidden behind the scattering medium.

[0067] Specifically, such as Figure 2 As shown, Figure 2This invention provides an imaging system for high-quality reconstruction of an unknown target image through a scattering medium. The system includes a laser source 1, a beam expander 2, a rotating frosted glass 3, a variable aperture 4, a spatial light modulator 5, a scattering medium 6, a variable aperture 7, a narrowband filter 8, and a camera 9, arranged sequentially along the optical axis. The laser source 1 outputs a stable light source; the beam expander 2 collimates the light generated by the laser source 1; the rotating frosted glass 3 modulates the light output by the laser source 1 into pseudo-thermal light; the variable aperture 4 constrains and collimates the pseudo-thermal light; the spatial light modulator 5 loads the target image under computer control; the scattering medium 6 is frosted glass; the variable aperture 7 collimates the scattered light; the narrowband filter 8 effectively filters out stray light from the optical path, and the camera 9 is a high-frame-rate, high-resolution CMOS camera.

[0068] This imaging system can continuously and rapidly acquire high-resolution speckle patterns containing a large amount of hidden target image information, and can significantly reduce the impact of external noise on image quality.

[0069] Specifically, in this embodiment, the center wavelength of the laser light source is 632.8nm, the scattering medium is frosted glass with a particle size of 220, the camera's frame rate is 165fps, and the resolution is 1920×1200.

[0070] Specifically, in this embodiment, the ratio of the distance u between the target and the scattering medium to the distance v between the scattering medium and the camera is controlled between 3.5 and 4, and the diameter of the variable aperture is 3 mm, in order to obtain the best imaging effect.

[0071] Specifically, in this embodiment, the target image output by the spatial light modulator in the S1 imaging system is as follows: Figure 3 As shown, the speckle pattern acquired using the imaging system is as follows: Figure 4 As shown.

[0072] S2. Segment and project the speckle pattern, and perform triple correlation operations;

[0073] Specifically, S2 includes:

[0074] S21. Set the segmentation interval and segmentation overlap rate, and segment the speckle pattern according to this standard, dividing the speckle image into n sub-speckle patterns. Adjacent sub-speckle patterns overlap to ensure they contain all the information of the image. The intensity I of the m-th sub-speckle pattern... m It is given by the following formula:

[0075] I m =O*S m

[0076] Among them, S m It is the point spread function at the center of the m-th sub-speckle.

[0077] Specifically, in this embodiment, the segmentation interval in S31 is set to 240, the segmentation overlap rate is set to 0.8, and the resolution is 3072×2048 to obtain 2009 sub-speckle data.

[0078] S22. All the sub-speckle patterns obtained by segmentation are uniformly projected onto the 0 to π area using Radon transform to obtain one-dimensional sub-speckle projection data.

[0079] Specifically, in this embodiment, the number of projections in S32 is set to 18, and each sub-speckle is transformed into 342×18 one-dimensional projection data through projection.

[0080] S23. Perform triple correlation on the one-dimensional sub-speckle data obtained from the projection. The measure of triple correlation is the bispectrum. The bispectrum obtained by triple correlation of the m-th sub-speckle projection can be expressed as:

[0081]

[0082] Where f1 and f2 are spatial frequencies, It is the position vector in the image. This indicates the conjugate operation.

[0083] S24. Extract the TC phase by calculating the bispectrum of each sub-speckle projection.

[0084] The TC phase is a bispectral phase spectrum, which is extracted by calculating the bispectral spectrum of each sub-speckle projection.

[0085] S3. Reconstruct the Fourier phase using a phase recursive algorithm guided by normalized weighting terms;

[0086] Specifically, S3 includes:

[0087] S31. Calculate the weighted average term W in the recursive process. m (i,p) can be represented as:

[0088]

[0089] in, It is the variance of the bispectral modulus, and the weighted average term W. m (i,p) can be viewed as the square of the bispectral signal-to-noise ratio.

[0090] S32. Introduce phasor difference D(f,Δf) and phase difference. As a constraint parameter in the recursive process, the phasor difference can be expressed as:

[0091]

[0092] S33. Introducing a weighted phasor Ω(i) to guide the normalized weighted terms in each iteration of the TC phase recursive Fourier phase algorithm, the weighted phasor at f = i in the q-th iteration of the m-th projection can be expressed as:

[0093]

[0094] Each iteration in the recursive process is performed by normalizing Ω(i), and the initial value of the iteration is the phase estimated by the recursive algorithm.

[0095] Understandably, in another embodiment, the initial value of the iterative operation in this step is zero.

[0096] S34. The weighted phasor Ω(i) is renormalized in each iteration, and the normalized weighted term R is obtained from the normalization operation. m (i) The condition for determining the iteration loop can be expressed as:

[0097]

[0098] When R m (i) The iteration process terminates when the change in continuous iterations is comparable to the noise.

[0099] The relationship between the S35 and TC phases and the target phase can be expressed as:

[0100]

[0101] The fourfold phase of the speckle TC phase β(f1,f2) is compared with the target Fourier phase. The relationship between them is recursively used in iterative calculations to obtain the target Fourier phase information.

[0102] S4. Repeat S3 to combine the recursively reconstructed Fourier phase with the Fourier amplitude extracted by speckle correlation to output the reconstructed target image.

[0103] Specifically, S4 includes:

[0104] S41. Repeat S3 to recursively reconstruct the target Fourier phase from the TC phase of all sub-speckle data, thus completing the Fourier phase reconstruction of the unknown target.

[0105] S42. Perform autocorrelation on the speckle pattern I(x). The calculation process can be expressed as follows:

[0106]

[0107] in The autocorrelation operation is represented by the autocorrelation function. The autocorrelation of the point spread function can be approximated as the impulse function within a certain range. Therefore, the autocorrelation of the speckle pattern can be approximately equivalent to the autocorrelation of the target image, and is expressed as:

[0108]

[0109] S43: According to Wiener-Khinchin's theorem, the Fourier amplitude of the target is calculated using the Fourier transform of the speckle map autocorrelation.

[0110]

[0111] S44. Combine the Fourier phase obtained by recursion with the Fourier amplitude obtained by autocorrelation to output the reconstructed target image.

[0112] Specifically, in this embodiment, the reconstructed target image output by S44 is as follows: Figure 5 As shown.

[0113] In summary, the scattering imaging method using normalized weighted terms to guide phase recursion successfully achieves efficient and high-quality reconstruction of unknown target image information hidden behind scattering media. The reconstructed image can completely preserve the original target's directional information, and has high imaging quality and fast computational efficiency. This method uses an imaging system that reconstructs high-quality unknown target images through scattering media to acquire high-quality speckle patterns for analysis and imaging. The method simplifies the triple correlation calculation process, effectively reducing redundant operations and significantly improving the algorithm's execution speed. Simultaneously, the introduction of triple correlation operations for phase reconstruction effectively preserves the directional information of the original target image, enabling high-precision reconstruction of unknown targets. Furthermore, the introduction of normalized weighted terms to guide recursion effectively enhances the algorithm's noise resistance and improves the convergence speed during iteration, effectively reducing the number of phase recovery iterations, thereby obtaining robust and noise-resistant imaging results.

[0114] The embodiments of the present invention specifically realize high-quality imaging of unknown targets hidden behind scattering media, and have broad practical application prospects in complex environments.

[0115] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions, and variations can be made to these embodiments, including components, without departing from the principles and spirit of the present invention, and these variations still fall within the protection scope of the present invention.

Claims

1. A scattering imaging method based on phase recursion guided by normalized weighting terms, characterized in that, Includes the following steps: S1. Obtain the speckle pattern of the unknown target hidden behind the scattering medium by using an imaging system that reconstructs a high-quality image of the unknown target through the scattering medium. S2. Segment and project the speckle pattern, and perform triple correlation operations; S3. Reconstruct the Fourier phase using a phase recursive algorithm guided by normalized weighting terms; Step S3 includes the following sub-steps: S31. Calculate the weighted average term W in the recursive process. m (i,p) is represented as: in, It is the variance of the bispectral modulus, and the weighted average term W. m (i,p) can be considered as the square of the bispectral signal-to-noise ratio; S32. Introduce phasor difference D(f,Δf) and phase difference. As a constraint parameter in the recursive process, the phasor difference is expressed as: S33. Introducing a weighted phasor Ω(i) to guide the normalized weighted terms in each iteration of the TC phase recursive Fourier phase algorithm, the weighted phasor at f = i in the q-th iteration of the m-th projection is represented as: Each iteration in the recursive process is performed by normalizing Ω(i), and the initial value of the iteration is the phase estimated by the recursive algorithm or is zero; S34. The weighted phasor Ω(i) is renormalized in each iteration, and the normalized weighted term R is obtained from the normalization operation. m (i) The condition for determining the iteration loop is expressed as: When R m (i) The iteration process terminates when the change in continuous iterations becomes equivalent to the noise; The relationship between the S35 and TC phases and the target phase is expressed as follows: The fourfold phase of the speckle TC phase β(f1,f2) is compared with the target Fourier phase. The relationship between them is recursively used in iterative calculations to obtain the target Fourier phase information; S4. Repeat S3 to combine the recursively reconstructed Fourier phase with the Fourier amplitude extracted by speckle correlation to output the reconstructed target image.

2. The scattering imaging method based on normalized weighted terms guided by phase recursion according to claim 1, characterized in that, In step S2, the method for speckle image segmentation and projection is as follows: The speckle pattern is segmented into n sub-speckle patterns by setting a segmentation interval and a segmentation overlap rate. Adjacent sub-speckle patterns overlap to ensure they contain all the information of the image. The intensity I of the m-th sub-speckle pattern is... m It is given by the following formula: I m =O*S m Among them, S m It is the point spread function at the center of the m-th sub-speckle.

3. The scattering imaging method based on phase recursion guided by normalized weighting terms according to claim 2, characterized in that, In step S2, the triple correlation calculation method is as follows: First, all the sub-speckle patterns obtained by segmentation are uniformly projected onto the 0 to π area using Radon transform to obtain one-dimensional sub-speckle projection data. Then, a triple correlation operation is performed on the one-dimensional sub-speckle data obtained by projection. The quantity that measures the triple correlation is the bispectrum. The bispectrum obtained by triple correlation of the m-th sub-speckle projection is represented as follows: Where f1 and f2 are spatial frequencies, It is a position vector in the image. Indicates conjugate operation; Finally, after calculating the bispectral spectrum of each sub-speckle projection, the TC phase is extracted as the phase spectrum of the bispectral spectrum.

4. The scattering imaging method based on normalized weighted terms guided by phase recursion according to claim 3, characterized in that, In step S33, each iteration of the recursive process is performed by normalizing Ω(i), and the initial value of the iteration is the phase estimated by the recursive algorithm.

5. The scattering imaging method based on normalized weighted term-guided phase recursion according to claim 3, characterized in that, In step S33, each iteration of the recursive process is performed by normalizing Ω(i), and the initial value of the iteration is 0.

6. The scattering imaging method based on phase recursion guided by normalized weighting terms according to claim 3, characterized in that, In step S34, a normalized weighting term R is introduced. m (i) Guide the iterative process in the recursive operation; the termination condition of the iteration is R. m (i) The iteration process terminates when the change in continuous iterations is comparable to the noise.

7. The scattering imaging method based on phase recursion guided by normalized weighting terms according to claim 3, characterized in that, Step S4 includes the following sub-steps: S41. Repeat S3 to recursively reconstruct the target Fourier phase from the TC phase of all sub-speckle data. S42. Perform autocorrelation on the speckle pattern I(x). The calculation process is as follows: in The autocorrelation operation is represented by the autocorrelation function. The autocorrelation of the point spread function is approximately equivalent to the impulse function within a certain range. Therefore, the autocorrelation of the speckle pattern is approximately equivalent to the autocorrelation of the target image, and is represented as: S43. According to Wiener-Khinchin's theorem, the Fourier amplitude of the target is calculated using the Fourier transform of the speckle pattern autocorrelation: S44. Combine the Fourier phase obtained by recursion with the Fourier amplitude obtained by autocorrelation to output the reconstructed target image.

Images