A diffraction imaging system and a method for quickly restoring an image

Abstract

The present invention discloses a diffraction imaging system and a rapid image restoration method. Addressing the problem of non-imaging-order light becoming strong background noise due to the multi-order diffraction characteristics of diffraction imaging elements, a rapid image restoration method based on an equivalent imaging model of the diffraction imaging element is proposed. The diffraction imaging system comprises a monochromatic extended light source, an imaging target, a far-field lens, a Fresnel lens, a beam reduction lens, and an imaging detection camera. By using an optimization method and a Wiener filter, the present invention solves the equivalent point spread function of the Fresnel lens. This method then achieves rapid image restoration of the imaging system based on the imaging equivalent model, improving the imaging quality of the Fresnel lens.

Description

Technical Field

[0001] The present invention belongs to the technical field of thin film diffraction imaging, and in particular relates to a diffraction imaging system and a method for quickly restoring an image. Background Art

[0002] While thin-film diffraction imaging offers advantages such as ultra-large aperture, lightweight, loose tolerances, and ease of unfolding and folding, it also faces the dual problems of low imaging signal-to-noise ratio due to multiple diffraction orders and reduced image resolution due to wavefront aberrations. Non-imaging order diffracted light becomes strong background noise, causing a severe drop in the signal-to-noise ratio while also restricting the ability to improve the resolution of the imaging system through wavefront aberration correction. The main imaging elements of thin-film diffraction imaging include Fresnel zone plates (FZP), photon sieves, composite zone plates, and spliced ​​zone plates. Fresnel zone plates, with their ultra-lightness, thinness, and foldability, have become a staple of large-aperture telescopes in space.

[0003] FZP is a series of concentric micro-nanostructured rings that are etched. These rings change the phase of the incident light wavefront, allowing the light to converge at the main focus for imaging. Since the continuous phase comb-shaped rings of FZP are difficult to process, a step-by-step approach is actually used to approximate the continuous structure. The relationship between the ideal diffraction efficiency and the number of steps is:

[0004]

[0005] Where N is the number of steps in the FZP ring structure, λ0 is the design wavelength, λ is the operating wavelength, and m is the diffraction order.

[0006] On the one hand, due to technical limitations, ring-shaped structural defects are inevitable. The actual diffraction efficiency of the Nier zone plate is much lower than the theoretical value (40.5% for the second step and 81% for the fourth step). The non-imaging diffraction-level light in the FZP imaging system will become a strong background noise for the main imaging-level light, causing a serious decrease in the system signal-to-noise ratio. The specific manifestation is a thick layer of foggy gray covering the imaging target image.

[0007] On the other hand, due to annular defects, material deformation, and beam propagation through inhomogeneous media, the imaging beam wavefront is severely distorted, resulting in a sharp degradation in the imaging resolution of the diffraction imaging system. Improving the imaging resolution by correcting the wavefront distortion is limited by the low signal-to-noise ratio of the imaging system. The model of diffraction imaging caused by the coupling of multiple diffraction orders and wavefront aberrations can be equivalent to:

[0008]

[0009] Where I is the actual imaging of the target, I0 is the ideal imaging of the target, η0 is the diffraction efficiency of the imaging order, PSF0 and PSF mare the point spread functions of the primary and other order lights respectively, and N is the noise.

[0010] Diffraction imaging is affected by the dual effects of multi-order diffraction and wavefront distortion, and the imaging quality degrades sharply. Therefore, this application proposes a fast image restoration method based on the model of formula (2) to address the problem of low signal-to-noise ratio of multi-order diffraction. Summary of the Invention

[0011] In order to solve the technical problems existing in the background technology, the present invention aims to provide a diffraction imaging system and a rapid image restoration method, which is suitable for image post-processing and restoration of thin film diffraction imaging systems such as Fresnel zone plates, photon sieves, and spliced ​​diffraction telescopes. Through the optimization method and Wiener filter, the equivalent point spread function of the Fresnel zone plate is solved, thereby realizing rapid restoration of the imaging system image according to the imaging equivalent model and improving the imaging quality of the Fresnel lens.

[0012] In order to solve the technical problem, the technical solution of the present invention is:

[0013] A diffraction imaging system includes: a monochromatic light expansion light source, an imaging target, a far-field lens, a diffraction imaging element, a beam reduction lens, an imaging detection camera, and a control computer;

[0014] The imaging target, far-field lens, diffraction imaging element, beam reduction lens and imaging detection camera are sequentially arranged and distributed along the optical axis of the monochromatic light expansion light source, and the imaging detection camera signal is connected to the control computer.

[0015] Furthermore, the diffraction imaging element adopts a Fresnel zone plate.

[0016] Furthermore, the light beam emitted by the monochromatic extended light source passes through the imaging target and the far-field lens to the Fresnel zone plate, and then is focused by the focusing lens and then enters the imaging detection camera for imaging.

[0017] Furthermore, the control computer is preset with a restoration calculation unit for the image of the diffraction imaging system. The restoration calculation unit for the image of the diffraction imaging system obtains image data according to the imaging detection camera, measures the primary diffraction light efficiency and the corresponding point spread function, uses an optimization algorithm to solve the optimal diffraction efficiency and the equivalent point spread function of the non-imaging light, and finally uses a Wiener filter to quickly realize the rapid restoration of the imaging system image.

[0018] A method for restoring an image of a diffraction imaging system, applied to any of the above-mentioned diffraction imaging systems, the method comprising:

[0019] Step S1: Replace the light source and imaging target with a point light source, and obtain the far-field spot of the imaging-order light of the Fresnel zone plate through the imaging camera, that is, the primary-order PSF0 of the Fresnel zone plate;

[0020] Step S2: Initialize the point spread function PSF corresponding to the non-imaging level light of the Fresnel zone plate m and the primary-order light diffraction efficiency η0, and bring it into the Fresnel imaging model;

[0021]

[0022] Where I is the image of the Fresnel imaging system detected by the camera, I0 is the imaging target, N is the imaging noise, Represents convolution. According to the imaging model, the ideal image of the target is solved as:

[0023] I0=deconvwr(IN,η0*PSF0+(1-η0)*PSF m ) (2)

[0024] Where deconvwr is deconvolution. The actual image I, the primary point spread function PSF0 and the corresponding diffraction efficiency η0 can be obtained by the Fresnel imaging system. The PSF is obtained by blind optimization. m and the correction value of the diffraction efficiency η0;

[0025] Step S3: Randomly generate a △PSF that follows a Bernoulli distribution m and △η0, a two-dimensional Wiener filter is used to implement the deconvolution of formula (2):

[0026] I 0+ =deconvwr(I,(η0+△η0)*PSF0+(1-η0-△η0)*(PSF m +△PSF m ),SNR) (3)

[0027] Where SNR is the signal-to-noise ratio;

[0028] Step S4: Calculate image I 0+ The corresponding clarity function J + :

[0029] J + =sum(sum(I (x,y)0+ -mean(I (x,y)0+ ) 2 ) / (x*y) (4)

[0030] In the formula, sum is the sum function, mean is the mean function, I (x,y)0+ For image I 0+ The coordinates (x, y) correspond to the grayscale of the pixel;

[0031] Step S5: Use the same △PSF in the two-dimensional Wiener filterm , △η0, to achieve:

[0032] I 0- =deconvwr(I,(η0-△η0)*PSF0+(1-η0+△η0)*(PSF m -△PSF m ),SNR)(5)

[0033] Step S6: Calculate image I 0- The corresponding clarity function J - :

[0034] J - =sum(sum(I (x,y)0- -mean(I (x,y)0- ) 2 ) / (x*y) (6)

[0035] Step S7: Implement PSF using a parallel gradient stochastic algorithm m And iterative update of diffraction efficiency η0:

[0036] PSF m =PSF m +gama1*(J + -J - )*△PSF m (7)

[0037] η0=η0+gama2*(J + -J - )*△η0 (8)

[0038] Where gama1 and gama2 are the gain steps of iterative optimization;

[0039] Step S8: Continuously iterate steps S3 to S7 until the image clarity J converges, and solve the equivalent model of the Fresnel zone plate imaging system as shown in formula (1) to obtain the PSF m and the optimal solution of diffraction efficiency η0;

[0040] Step S9: Replace any imaging target and use the Wiener filter to quickly restore the Fresnel imaging system image I, that is, obtain the imaging target image I0:

[0041] I0=deconvwr(IN,η0*PSF0+(1-η0)*PSF m , SNR) (9).

[0042] Compared with the prior art, the advantages of the present invention are:

[0043] The present invention provides a diffraction imaging system and a rapid image restoration method, which can be applied to the rapid restoration of images of diffraction imaging systems such as Fresnel zone plates, photon sieves, and thin-film splicing telescopes. The primary diffraction light efficiency and the corresponding point spread function can be experimentally measured for different systems. The optimal diffraction efficiency and the equivalent point spread function of non-imaging light can be solved through an optimization algorithm. Finally, a Wiener filter is used to quickly realize the rapid restoration of the imaging system image, laying the foundation for subsequent system wavefront correction and image post-processing, and ultimately improving the imaging quality of the diffraction imaging system. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a schematic diagram of the experimental device of a diffraction imaging system of the present invention;

[0045] Figure 2 is the point spread function corresponding to the primary-order diffraction light of the Fresnel zone plate imaging system;

[0046] Figure 3 It is the optimal convergence curve of the equivalent point spread function corresponding to the non-imaging diffracted light of the Fresnel zone plate imaging system;

[0047] Figure 4 is the equivalent point spread function corresponding to the non-imaging diffracted light of the Fresnel zone plate imaging system;

[0048] Figure 5 This is the actual imaging image of the Fresnel zone plate imaging system;

[0049] Figure 6 This is a flowchart of an iterative optimization method for solving the equivalent point spread function of a model-based fast image restoration method for a diffraction imaging system;

[0050] Figure 7 It is a model-based fast image restoration method.

[0051] Reference numerals:

[0052] 1- Monochromatic light expansion light source; 2- Imaging target; 3- Far-field lens; 4- Diffraction imaging element; 5- Beam reduction lens; 6- Imaging detection camera; 7- Control computer. DETAILED DESCRIPTION

[0053] The specific implementation of the present invention is described below in conjunction with examples:

[0054] It should be noted that the structures, proportions, sizes, etc. shown in this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the conditions under which the present invention can be implemented. Any structural modification, change in proportional relationship or adjustment of size should still fall within the scope of the technical content disclosed in the present invention without affecting the efficacy and purpose that can be achieved by the present invention.

[0055] At the same time, the terms such as "upper", "lower", "left", "right", "middle" and "one" quoted in this specification are only for the convenience of description and are not used to limit the scope of implementation of the present invention. Changes or adjustments to their relative relationships should be regarded as the scope of implementation of the present invention without substantially changing the technical content.

[0056] Example 1:

[0057] This embodiment solves the problem that the diffraction imaging system is affected by the multi-order diffraction of the diffraction imaging element, resulting in a sharp degradation of the image signal-to-noise ratio and a decrease in image quality. At the same time, it further restricts the correction of the wavefront distortion and the improvement of the resolution of the imaging system. This embodiment designs and provides a fast image restoration method based on an imaging model. It uses a monochromatic light source, a far-field lens, a Fresnel zone plate, a beam reduction lens, a CCD camera and other imaging optical paths and a computer control system to implement an imaging Fresnel zone plate imaging system, providing the prerequisites for subsequent improvement of image restoration in the imaging experiment process, such as Figure 1-7 shown.

[0058] like Figure 1 As shown in the figure, in the experimental imaging system, a monochromatic light source illuminates the imaging object, and the light beam passes through the far-field lens, making the imaging target equivalent to infinity. The imaging beam is focused by the Fresnel zone plate, and the beam reduction lens adjusts the light beam passing through the Fresnel zone plate to adapt to the target surface of the imaging camera. The camera collects the image and uploads it to the control computer via Ethernet for image restoration.

[0059] The specific implementation steps are as follows:

[0060] Step 1: Replace the light source and imaging target with a point light source, and use the imaging camera to obtain the far-field spot of the imaging-order light of the Fresnel zone plate, that is, the primary-order PSF0 of the Fresnel zone plate;

[0061] Step 2: Initialize the point spread function (PSF) corresponding to the non-imaging light of the Fresnel zone plate m and the primary-order light diffraction efficiency η0, and bring it into the Fresnel imaging model;

[0062]

[0063] Where I is the image of the Fresnel imaging system detected by the camera, I0 is the imaging target, N is the imaging noise, Represents convolution. According to the imaging model, the ideal image of the target is solved as:

[0064] I0=deconvwr(IN,η0*PSF0+(1-η0)*PSF m ) (2)

[0065] Where deconvwr is deconvolution. The actual image I, the primary point spread function PSF0 and the corresponding diffraction efficiency η0 can be obtained by the Fresnel imaging system. The PSF is obtained by blind optimization. m and the correction value of the diffraction efficiency η0;

[0066] Step 3: Randomly generate a △PSF that follows a Bernoulli distribution m and △η0, a two-dimensional Wiener filter is used to implement the deconvolution of formula (2).

[0067] I 0+ =deconvwr(I,(η0+△η0)*PSF0+(1-η0-△η0)*(PSF m +△PSF m ),SNR) (3)

[0068] Where SNR is the signal-to-noise ratio;

[0069] Step 4: Calculate image I 0+ The corresponding clarity function J + :

[0070] J + =sum(sum(I (x,y)0+ -mean(I (x,y)0+ ) 2 ) / (x*y) (4)

[0071] In the formula, sum is the sum function, mean is the mean function, I (x,y)0+ For image I 0+ The coordinates (x, y) correspond to the grayscale of the pixel;

[0072] Step 5: Apply the same ∆PSF to the 2D Wiener filter m , △η0, to achieve:

[0073] I 0- =deconvwr(I,(η0-△η0)*PSF0+(1-η0+△η0)*(PSF m -△PSF m ),SNR) (5)

[0075] Step 6: Calculate image I 0- The corresponding clarity function J - :

[0076] J - =sum(sum(I (x,y)0- -mean(I (x,y)0- ) 2 ) / (x*y) (6)

[0077] Step 7: Implement PSF using a parallel gradient stochastic algorithm m and iterative update of the diffraction efficiency η0;

[0078] PSF m =PSF m +gama1*(J + -J - )*△PSF m (7)

[0079] η0=η0+gama2*(J + -J - )*△η0 (8)

[0080] Where gama1 and gama2 are the gain steps of iterative optimization;

[0081] Step 8: Continuously iterate steps 3 to 7 until the image clarity J converges, and solve the equivalent model of the Fresnel zone plate imaging system as shown in equation (1) to obtain the PSF m and the optimal solution of diffraction efficiency η0;

[0082] Step 9: Replace any imaging target and use the Wiener filter to quickly restore the Fresnel imaging system image I, that is, obtain the imaging target image I0.

[0083] I0=deconvwr(IN,η0*PSF0+(1-η0)*PSF m ,SNR) (9)

[0084] In summary, the present invention implements a model-based rapid image restoration method for a diffraction imaging system. A light beam emitted by a monochromatic extended light source 1 passes through an imaging target 2 and a far-field lens 3 before reaching a Fresnel zone plate 4. The beam is then reduced by a beam reduction lens 5 before entering an imaging detection camera 6 for imaging. By using an optimization method and a Wiener filter, the equivalent point spread function of the Fresnel zone plate is determined. This allows for rapid image restoration of the imaging system based on the imaging equivalent model, improving the imaging quality of the Fresnel lens.

[0085] The preferred embodiments of the present invention are described in detail above, but the present invention is not limited to the above embodiments. Various changes can be made within the knowledge of ordinary technicians in this field without departing from the scope of the present invention.

[0086] Many other changes and modifications can be made without departing from the spirit and scope of the present invention. It should be understood that the present invention is not limited to the specific embodiments, and the scope of the present invention is defined by the appended claims.

Claims

1. A diffraction imaging system, characterized in that: include: A monochromatic light expansion light source, an imaging target, a far-field lens, a diffractive imaging element, a beam reduction lens, an imaging detection camera, and a control computer; the imaging target, far-field lens, diffractive imaging element, beam reduction lens, and imaging detection camera are sequentially arranged and distributed along the optical axis of the monochromatic light expansion light source, and the imaging detection camera signal is connected to the control computer; the diffractive imaging element uses a Fresnel zone plate; the light beam emitted by the monochromatic light expansion light source passes through the imaging target and the far-field lens before reaching the Fresnel zone plate, and then is beam-reduced by the beam reduction lens before entering the imaging detection camera for imaging; the control computer is pre-set with a diffractive imaging system image restoration calculation unit, which executes an image restoration method: acquiring image data from the imaging detection camera, measuring the primary diffraction light efficiency and the corresponding point spread function, using an optimization algorithm to solve for the optimal diffraction efficiency and the equivalent point spread function of the non-imaging light, and finally restoring the imaging system image using a Wiener filter; the image restoration method specifically includes: Step S1: Replace the light source and imaging target with a point light source, and obtain the far-field spot of the imaging-order light of the Fresnel zone plate through the imaging camera, that is, the primary-order PSF0 of the Fresnel zone plate; Step S2: Initialize the point spread function PSF corresponding to the non-imaging level light of the Fresnel zone plate m and the primary-order light diffraction efficiency η0, and bring it into the Fresnel imaging model; Where I is the image of the Fresnel imaging system detected by the camera, I0 is the imaging target, N is the imaging noise, Represents convolution. According to the imaging model, the ideal image of the target is solved as: I0=deconvwr(I-N,η0*PSF0+(1-η0)*PSF m ) (2) Where deconvwr is deconvolution. The actual image I, the primary point spread function PSF0 and the corresponding diffraction efficiency η0 can be obtained by the Fresnel imaging system. The PSF is obtained by blind optimization. m and the correction value of the diffraction efficiency η0; Step S3: Randomly generate a △PSF that follows a Bernoulli distribution m and △η0, a two-dimensional Wiener filter is used to implement the deconvolution of formula (2): I 0+ =deconvwr(I,(η0+△η0)*PSF0+(1-η0-△η0)*(PSF m +△PSF m ),SNR) (3) Where SNR is the signal-to-noise ratio; Step S4: Calculate image I 0+ The corresponding clarity function J + : J + =sum(sum(I (x,y)0+ -mean(I (x,y)0+ ) 2 ) / (x*y) (4) In the formula, sum is the sum function, mean is the mean function, I (x,y)0+ For image I 0+ The coordinates (x, y) correspond to the grayscale of the pixel; Step S5: Using the same △PSF in the two-dimensional Wiener filter m , △η0, to achieve: I 0- =deconvwr(I,(η0-△η0)*PSF0+(1-η0+△η0)*(PSF m -△PSF m ),SNR) (5) Step S6: Calculate image I 0- The corresponding clarity function J - : J - =sum(sum(I (x,y)0- -mean(I (x,y)0- ) 2 ) / (x*y) (6) Step S7: Implement PSF using a parallel gradient stochastic algorithm m And iterative update of diffraction efficiency η0: PSF m =PSF m +gama1*(J + -J - )*△PSF m (7) η0=η0+gama2*(J + -J - )*△η0 (8) Where gama1 and gama2 are the gain steps of iterative optimization; Step S8: Continuously iterate steps S3 to S7 until the image clarity J converges, and solve the equivalent model of the Fresnel zone plate imaging system as shown in formula (1) to obtain the PSF m and the optimal solution of diffraction efficiency η0; Step S9: Replace any imaging target and use the Wiener filter to restore the Fresnel imaging system image I, that is, obtain the imaging target image I0: I0=deconvwr(I-N,η0*PSF0+(1-η0)*PSF m ,SNR) (9)。

Images