A method for evaluating reconstructed image quality based on sparse reconstruction model

Abstract

The present invention belongs to the technical field of photon integrated interference imaging, and in particular, relates to a method for evaluating the quality of a reconstructed image based on a sparse reconstruction model. The method comprises: S1: calculating the frequency sampling points of each paired baseline; S2: calculating the size of the reconstructed image; S3: adjusting the size of the original image; S4: generating a frequency domain mask based on the frequency sampling points of each paired baseline obtained in step S1; S5: updating the reconstructed image spectrum using the frequency domain mask; S6: performing Fourier transform on the updated reconstructed image spectrum to obtain a reconstructed image; S7: calculating the lateral gradient and longitudinal gradient of the reconstructed image; S8: calculating the lateral gradient and longitudinal gradient of the original image; S9: calculating the gradient root mean square error between the reconstructed image and the original image; S10: evaluating the image quality of the reconstructed image based on the gradient root mean square error. The present invention provides powerful guidance for system design and reconstructed image quality.

Description

Technical Field

[0001] The present invention belongs to the technical field of photon integrated interference imaging, and in particular relates to a reconstructed image quality evaluation method based on a sparse reconstruction model. Background Art

[0002] In the field of astronomical observation and remote sensing imaging, in order to improve the imaging resolution, it is usually chosen to increase the aperture of the optical system. However, as the aperture of the optical system increases, its manufacturing cost also increases exponentially. Photon integrated interferometric imaging technology provides a new technical approach to replace traditional large-aperture optical telescopes. The use of photon integrated chips (PICs) to realize the beam synthesis of the interference array can significantly improve the real-time, sensitivity, and scalability of the optical system, and effectively enhance the stability and reliability of the system. The processing and manufacturing are simple, the process is stable, and the manufacturing cost is effectively reduced. However, unlike the imaging mode in the traditional Cartesian coordinate system, the frequency sampling method in the current photon integrated interferometric imaging is polar coordinate discrete sampling. Discrete sampling will result in a large number of frequency loss, resulting in a large number of artifacts in the reconstructed image. When evaluating the quality of reconstructed images with artifacts, traditional evaluation methods, such as root mean square error and structural similarity, cannot well reflect the quality of the reconstructed image. Summary of the invention

[0003] In view of this, the invention aims to provide a reconstructed image quality evaluation method based on a sparse reconstruction model to solve the problem that the reconstructed image quality evaluation method used in the prior art has certain limitations and cannot correctly evaluate the quality of the reconstructed image. The invention is based on the similarity of the gradients of the reconstructed image and the original image, calculates the root mean square error of the gradient image, can effectively extract the effective features in the reconstructed image and compare them, and reduce the impact of artifacts on the quality evaluation of the reconstructed image.

[0004] To achieve the above object, the technical solution created by the present invention is implemented as follows:

[0005] A reconstructed image quality evaluation method based on a sparse reconstruction model specifically comprises the following steps:

[0006] S1: Calculate the frequency sampling points of each paired baseline according to the baseline configuration and baseline pairing method of the photon integrated interferometric imaging system;

[0007] S2: Calculate the size (m, m) of the reconstructed image based on the frequency sampling points of each paired baseline;

[0008] S3: resize the original image to (m, m);

[0009] S4: Generate a frequency domain mask based on the frequency sampling points of each paired baseline obtained in step S1;

[0010] S5: Perform Fourier transform on the original image to obtain a reconstructed image spectrum, and update the reconstructed image spectrum using a frequency domain mask;

[0011] S6: performing inverse Fourier transform on the updated reconstructed image spectrum to obtain a reconstructed image;

[0012] S7: Calculate the horizontal gradient and the vertical gradient of the reconstructed image;

[0013] S8: Calculate the horizontal gradient and the vertical gradient of the original image;

[0014] S9: Calculate the root mean square error of the gradient between the reconstructed image and the original image;

[0015] S10: Evaluate the image quality of the reconstructed image based on the gradient root mean square error.

[0016] Furthermore, in step S1, the frequency sampling points of each paired baseline are calculated as follows:

[0017] ;

[0018] Among them, B p is the paired baseline of group p, is the jth imaging wavelength of the photon integrated interferometric imaging system, z is the imaging distance of the photon integrated interferometric imaging system, is the frequency sampling point of the photon integrated interferometric imaging system, and there are n groups of frequency sampling points in total, ( x , y ) is the spatial coordinate, ( , ) is the spatial coordinate of the p-th group of paired baselines.

[0019] Furthermore, in step S2, the formula for calculating the size m of the reconstructed image is:

[0020] ;

[0021] ;

[0022] ;

[0023] ;

[0024] Among them, R is the number rounded up, is the ratio of the maximum frequency to the fundamental frequency of the photon integrated interferometric imaging system, is the maximum frequency of the photon integrated interferometric imaging system, is the fundamental frequency of the photon integrated interferometric imaging system, z is the imaging distance of the photon integrated interferometric imaging system, B minis the shortest paired baseline, B max is the longest paired baseline, is the minimum imaging wavelength of the photon integrated interferometric imaging system, is the maximum imaging wavelength of the photon integrated interferometric imaging system.

[0025] Furthermore, in step S4, the expression of the frequency domain mask is:

[0026] ;

[0027] Where P is the frequency domain mask, is the frequency sampling point of the photon integrated interferometric imaging system, and there are n groups of frequency sampling points in total. is the nth group of frequency sampling points.

[0028] Furthermore, in step S5, the formula used to update the spectrum using the frequency domain mask is:

[0029] ;

[0030] Among them, S is the updated reconstructed image spectrum, is the coordinate of the nth group of frequency sampling points, To expand the horizontal coordinate of the nth group of frequency sampling points by N times, To expand the vertical coordinates of the nth group of frequency sampling points by N times, is the complex visibility of the nth group of frequency sampling points.

[0031] Furthermore, in step S6, the calculation formula for reconstructing the image is:

[0032] ;

[0033] Where IFT is the inverse Fourier transform, (ζ,η) is the reconstructed image The coordinates of To reconstruct the image.

[0034] Furthermore, in step S7, the calculation formula used to calculate the transverse gradient and the longitudinal gradient of the reconstructed image is:

[0035] ;

[0036] ;

[0037] in, To reconstruct the image The lateral gradient of To reconstruct the image The vertical gradient.

[0038] Furthermore, in step S8, the calculation formula used to calculate the horizontal gradient and the vertical gradient of the original image is:

[0039] ;

[0040] ;

[0041] in, For the original image The lateral gradient of For the original image The vertical gradient.

[0042] Furthermore, in step S9, the root mean square error of the gradient between the reconstructed image and the original image is calculated. The calculation formula used is:

[0043] .

[0044] Furthermore, the larger the gradient root mean square error value, the worse the image quality of the reconstructed image, and the smaller the gradient root mean square error value, the better the image quality of the reconstructed image.

[0045] Compared with the prior art, the invention can achieve the following beneficial effects:

[0046] (1) The present invention creates a reconstructed image quality evaluation method based on a sparse reconstruction model, which can effectively overcome the limitations of traditional evaluation methods by performing MSE comparison between the gradient of the reconstructed image and the gradient of the original image, and can provide a powerful criterion for the array arrangement design of photon integrated interferometric imaging and the quality evaluation of the reconstructed image.

[0047] (2) The reconstructed image quality evaluation method based on the sparse reconstruction model created by the present invention can overcome the problem of a large number of artifacts in the reconstructed image caused by non-uniform sparse sampling, effectively reduce the artifacts of the reconstructed image, and improve the signal-to-noise ratio of the reconstructed image. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The drawings constituting part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments and descriptions of the present invention are used to explain the present invention and do not constitute an improper limitation on the present invention. In the drawings:

[0049] Figure 1 A schematic diagram of a flow chart of a reconstructed image quality evaluation method based on a sparse reconstruction model according to an embodiment of the present invention;

[0050] FIG. 2 (a) is a schematic diagram showing the distribution of frequency sampling points of the ESPIA scheme according to an embodiment of the present invention;

[0051] FIG2( b ) is a schematic diagram showing the distribution of frequency sampling points of the HSLMA scheme according to an embodiment of the present invention;

[0052] FIG2( c ) is a schematic diagram showing the distribution of frequency sampling points of the SPIDER solution according to an embodiment of the present invention;

[0053] Figure 3 Creating the original image described in the embodiment of the present invention;

[0054] FIG4 (a) is a schematic diagram of a reconstructed image obtained based on the ESPIA scheme according to an embodiment of the present invention;

[0055] FIG4( b ) is a schematic diagram of a reconstructed image obtained based on the HSLMA scheme according to an embodiment of the present invention;

[0056] FIG4( c ) is a schematic diagram of a reconstructed image obtained based on the SPIDER scheme according to an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the invention more clear, the invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described here are only used to explain the invention and do not constitute a limitation of the invention.

[0058] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0059] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", etc. are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", etc. may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0060] In the description of the invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installation", "connection" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the invention can be understood according to specific circumstances.

[0061] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments.

[0062] like Figure 1 As shown, the present invention proposes a reconstructed image quality evaluation method based on a sparse reconstruction model, which specifically includes the following steps:

[0063] S1: Calculate the frequency sampling points of each paired baseline according to the baseline configuration and baseline pairing method of the photon integrated interferometric imaging system.

[0064] The frequency sampling points of each paired baseline are calculated as follows:

[0065] ;

[0066] Among them, B p is the paired baseline of group p, is the jth imaging wavelength of the photon integrated interferometric imaging system, z is the imaging distance of the photon integrated interferometric imaging system, is the frequency sampling point of the photon integrated interferometric imaging system, and there are n groups of frequency sampling points in total, ( x , y ) is the spatial coordinate, ( , ) is the spatial coordinate of the p-th group of paired baselines.

[0067] S2: Calculate the size (m, m) of the reconstructed image based on the frequency sampling points of each paired baseline.

[0068] The formula for calculating the size m of the reconstructed image is:

[0069] ;

[0070] ;

[0071] ;

[0072] ;

[0073] Among them, R is the number rounded up, is the ratio of the maximum frequency to the fundamental frequency of the photon integrated interferometric imaging system, is the maximum frequency of the photon integrated interferometric imaging system, is the fundamental frequency of the photon integrated interferometric imaging system, z is the imaging distance of the photon integrated interferometric imaging system, B min is the shortest paired baseline, B max is the longest paired baseline, is the minimum imaging wavelength of the photon integrated interferometric imaging system, is the maximum imaging wavelength of the photon integrated interferometric imaging system.

[0074] S3: Resize the original image to (m, m).

[0075] S4: Generate a frequency domain mask based on the frequency sampling points of each paired baseline obtained in step S1.

[0076] The expression of the frequency domain mask is:

[0077] ;

[0078] Where P is the frequency domain mask, are the frequency sampling points of the photon integrated interferometric imaging system, and there are n groups of frequency sampling points in total.

[0079] S5: Perform Fourier transform on the original image to obtain a reconstructed image spectrum, and update the reconstructed image spectrum using a frequency domain mask.

[0080] ;

[0081] Among them, S is the updated reconstructed image spectrum, is the coordinate of the nth group of frequency sampling points, To expand the horizontal coordinate of the nth group of frequency sampling points by N times, To expand the vertical coordinates of the nth group of frequency sampling points by N times, is the complex visibility of the nth group of frequency sampling points.

[0082] S6: Perform inverse Fourier transform on the updated reconstructed image spectrum to obtain a reconstructed image.

[0083] The calculation formula for reconstructing the image is:

[0084] ;

[0085] Where IFT is the inverse Fourier transform, (ζ,η) is the reconstructed image The coordinates of To reconstruct the image.

[0086] S7: Calculate the horizontal gradient and the vertical gradient of the reconstructed image.

[0087] The calculation formula used to calculate the horizontal gradient and vertical gradient of the reconstructed image is:

[0088] ;

[0089] ;

[0090] in, To reconstruct the image The lateral gradient of To reconstruct the image The vertical gradient.

[0091] S8: Calculate the horizontal gradient and the vertical gradient of the original image.

[0092] The calculation formula used to calculate the horizontal gradient and vertical gradient of the original image is:

[0093] ;

[0094] ;

[0095] in, For the original image The lateral gradient of For the original image The vertical gradient.

[0096] S9: Calculate the root mean square error of the gradient of the reconstructed image and the original image.

[0097] Calculate the root mean square error of the gradient between the reconstructed image and the original image The calculation formula used is:

[0098] .

[0099] S10: Evaluate the image quality of the reconstructed image based on the gradient root mean square error.

[0100] The larger the gradient root mean square error value, the worse the image quality of the reconstructed image, and the smaller the gradient root mean square error value, the better the image quality of the reconstructed image.

[0101] Figure 2 (a) to Figure 2 (c) show the frequency sampling point distribution diagrams of the three schemes respectively. Figure 3The original image shown in the figure is reconstructed to obtain the corresponding reconstructed images shown in Figures 4(a) to 4(c). The MSE evaluation method and the GRMSE evaluation method of the present invention are used to evaluate the image quality of the three reconstructed images. Table 1 shows the evaluation results of the image quality of the three reconstructed images using the MSE evaluation method. According to the image quality comparison results of the reconstructed images in Table 1, it can be clearly seen that the MSE evaluation value of the ESPIA scheme with clearer details is larger.

[0102] Table 1

[0103]

[0104] Table 2 shows the evaluation results of the image quality of the three reconstructed images using the GRMSE evaluation method. According to the image quality comparison results of the reconstructed images in Table 2, it can be clearly seen that using the image detail similarity as the evaluation basis can effectively describe the similarity between the reconstructed image and the original image in terms of image details and overcome the influence of artifacts. The GRMSE value of the ESPIA scheme is the smallest, which matches the image effect of the actual reconstructed image.

[0105] Table 2

[0106]

[0107] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the disclosure of the present invention can be performed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and this document does not limit this.

[0108] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A reconstructed image quality evaluation method based on a sparse reconstruction model, characterized in that: The specific steps include: S1: Calculate the frequency sampling points of each paired baseline according to the baseline configuration and baseline pairing method of the photon integrated interferometric imaging system; S2: Calculate the size (m, m) of the reconstructed image based on the frequency sampling points of each paired baseline; The formula for calculating the size m of the reconstructed image is: ； ； ； ； Among them, R is the number rounded up, is the ratio of the maximum frequency to the fundamental frequency of the photon integrated interferometric imaging system, is the maximum frequency of the photon integrated interferometric imaging system, is the fundamental frequency of the photon integrated interference imaging system, z is the imaging distance of the photon integrated interference imaging system, B min is the shortest paired baseline, B max is the longest paired baseline, is the minimum imaging wavelength of the photon integrated interferometric imaging system, is the maximum imaging wavelength of the photon integrated interference imaging system; S3: resize the original image to (m, m); S4: Generate a frequency domain mask based on the frequency sampling points of each paired baseline obtained in step S1; S5: Perform Fourier transform on the original image to obtain a reconstructed image spectrum, and update the reconstructed image spectrum using a frequency domain mask; S6: performing inverse Fourier transform on the updated reconstructed image spectrum to obtain a reconstructed image; S7: Calculate the horizontal gradient and the vertical gradient of the reconstructed image; S8: Calculate the horizontal gradient and the vertical gradient of the original image; S9: Calculate the root mean square error of the gradient between the reconstructed image and the original image; S10: Evaluate the image quality of the reconstructed image based on the gradient root mean square error.

2. The reconstructed image quality assessment method based on the sparse reconstruction model according to claim 1, characterized in that: In step S1, the frequency sampling points of each paired baseline are calculated as follows: ； Among them, B p is the paired baseline of group p, is the jth imaging wavelength of the photon integrated interference imaging system, z is the imaging distance of the photon integrated interference imaging system, are the frequency sampling points of the photon integrated interference imaging system, and there are n groups of frequency sampling points in total, ( x , y ) is the spatial coordinate, ( , ) is the spatial coordinate of the p-th group of paired baselines.

3. The reconstructed image quality assessment method based on the sparse reconstruction model according to claim 2, characterized in that: In step S4, the frequency domain mask is expressed as: ； Where P is the frequency domain mask, are the frequency sampling points of the photon integrated interference imaging system, and there are n groups of frequency sampling points in total, is the nth group of frequency sampling points.

4. The reconstructed image quality assessment method based on the sparse reconstruction model according to claim 3 is characterized in that: In step S5, the formula used to update the spectrum using the frequency domain mask is: ； Where S is the updated reconstructed image spectrum, is the coordinate of the nth group of frequency sampling points, To expand the horizontal coordinate of the nth group of frequency sampling points by N times, To expand the vertical coordinates of the nth group of frequency sampling points by N times, is the complex visibility of the nth group of frequency sampling points.

5. The reconstructed image quality assessment method based on the sparse reconstruction model according to claim 4 is characterized in that: In step S6, the calculation formula for reconstructing the image is: ； Wherein, IFT is the inverse Fourier transform, (ζ,η) is the reconstructed image The coordinates of To reconstruct the image.

6. The reconstructed image quality assessment method based on the sparse reconstruction model according to claim 5, characterized in that: In step S7, the calculation formula used to calculate the transverse gradient and the longitudinal gradient of the reconstructed image is: ； ； in, To reconstruct the image The lateral gradient of To reconstruct the image The vertical gradient.

7. The reconstructed image quality assessment method based on the sparse reconstruction model according to claim 6, characterized in that: In step S8, the calculation formula used to calculate the horizontal gradient and the vertical gradient of the original image is: ； ； in, For the original image The lateral gradient of For the original image The vertical gradient.

8. The reconstructed image quality assessment method based on the sparse reconstruction model according to claim 7, characterized in that: In step S9, the root mean square error of the gradient between the reconstructed image and the original image is calculated The calculation formula used is: 。 9. The reconstructed image quality assessment method based on the sparse reconstruction model according to claim 8, characterized in that: The larger the gradient root mean square error value, the worse the image quality of the reconstructed image, and the smaller the gradient root mean square error value, the better the image quality of the reconstructed image.

Images