Image quality evaluation method, system, equipment and medium for cubic phase mask imaging system

Abstract

A method, system, device, and medium for evaluating the image quality of a three-dimensional phase mask imaging system belong to the technical field of image quality evaluation methods for computational imaging systems and address the problem of low accuracy in image evaluation of three-dimensional phase mask imaging systems caused by existing evaluation methods. The method comprises obtaining a clear, uncoded image; obtaining a blurred, coded image; registering the clear, uncoded image and the blurred, coded image, obtaining the registered image and recording it as an image; processing the clear, uncoded image and recording it as an image; reconstructing the image, saving a region of the reconstructed image as an image, and determining, based on the temporary structural similarity between the recorded image and the image, the maximum structural similarity as the evaluation result of the three-dimensional phase mask coded imaging system.

Description

Technical Field

[0001] The present invention relates to the technical field of image quality evaluation methods for computational imaging systems, and in particular to an image quality evaluation method, system, equipment and medium for a cubic phase mask imaging system. Background Art

[0002] Cubic phase mask coded computational imaging technology is one of the current research hotspots in computational imaging technology. In traditional imaging systems, the performance of optical systems is highly dependent on precise optical component manufacturing and strict imaging environment control. Slight errors in optical components or slight changes in the imaging environment may lead to a significant decrease in imaging quality. Defocus of the optical system will make the image blurred, aberrations will cause image distortion, and environmental disturbances such as temperature changes and vibrations will affect the stability of optical components, thereby affecting the clarity and accuracy of the imaging. These unfavorable factors are difficult to completely avoid in many practical application scenarios, especially in fields with extremely high requirements for image clarity and accuracy, such as aerospace remote sensing, biomedical imaging, high-end industrial detection, etc. Therefore, it is particularly important to develop imaging technologies that can effectively deal with these unfavorable factors. The emergence of wavefront coded computational imaging technology provides a new approach to solving these problems. It introduces a specific phase mask on the pupil plane of the optical system to blur the image. (Point Spread Function) is shaped. This shaped It is insensitive to image quality degradation factors such as defocus and aberrations, allowing the imaging system to obtain relatively stable blurred images under different defocus amounts and certain aberrations. Subsequently, these blurred images are deconvolved and restored with the help of corresponding digital signal processing algorithms to reconstruct high-quality clear images. Compared with traditional imaging technologies, wavefront coding computational imaging technology is more robust in the imaging process. It does not require extremely precise adjustments to the optical system during imaging, which reduces the requirements of the optical system for manufacturing processes and assembly accuracy. It also reduces the impact of environmental factors on image quality, greatly expanding the scope of application of the imaging system.

[0003] However, current cubic phase mask coded computational imaging technology still faces several challenges. With the advancement of intelligent recognition and machine vision, subjective and no-reference evaluations alone cannot guarantee image data accuracy. Experiments with cubic phase mask coded computational imaging systems are limited by phase plate installation errors and the modulation of the cubic phase mask. Traditional image registration algorithms cannot fully align the blurred coded image with the clear uncoded image. This lacks the clear reference image and the blurred coded image to precisely align with it required for full-reference evaluation. Subjective and no-reference evaluation metrics are commonly used to evaluate cubic phase mask coded computational imaging systems. Furthermore, the restoration of the coded image can produce undesirable artifacts such as ringing and oversharpening. Ringing, typically manifesting as alternating bright and dark ring-like fringes near image edges, not only affects the visual quality but can also obscure important details, degrading image quality. Traditional subjective and no-reference evaluations cannot accurately reflect these deviations. In such circumstances, the continued use of traditional no-reference image quality assessment metrics will severely compromise their accuracy and reliability.

[0004] In summary, in the existing technology, subjective evaluation indicators and no-reference evaluation indicators are generally used to evaluate cubic phase mask coded computational imaging systems. However, the image sharpening and ringing effects in the restored coded image will lead to low accuracy of image evaluation. Summary of the Invention

[0005] The present invention solves the problem that the existing evaluation method leads to low accuracy of image evaluation of a cubic phase mask imaging system.

[0006] The image quality evaluation method of the three-dimensional phase mask imaging system of the present invention comprises the following steps:

[0007] Step S1, obtaining a clear and uncoded image;

[0008] Step S2, obtaining a fuzzy coded image;

[0009] Step S3: register the clear uncoded image and the fuzzy coded image, and obtain the registered fuzzy coded image as image ;

[0010] After processing the clear uncoded image, it is recorded as image ;

[0011] Step S4, image Reconstruct the reconstructed image of Save area as image , based on the recorded images With image Temporary structural similarity of , determine the maximum structural similarity Evaluation results of the cubic phase mask coded imaging system.

[0012] Furthermore, in the embodiment of the present invention, in step S1, obtaining a clear and uncoded image is specifically as follows:

[0013] After the cubic phase mask encoding plate is removed from the cubic phase mask encoding imaging system, the target scene is photographed to obtain a clear non-encoded image.

[0014] Furthermore, in the embodiment of the present invention, in step S2, the fuzzy coded image is obtained, specifically:

[0015] After the cubic phase mask encoding plate is installed in the cubic phase mask encoding imaging system, the target scene is photographed to obtain a fuzzy encoded image.

[0016] Furthermore, in the embodiment of the present invention, in step S3, the clear uncoded image is processed and recorded as image , specifically:

[0017] Select a clear area within the depth of focus in a clear uncoded image and obtain area, will be clear without coding image Region is recorded as image .

[0018] Furthermore, in the embodiment of the present invention, in the step S3, the clear uncoded image and the fuzzy coded image are registered, and the fuzzy coded image after registration is obtained and recorded as image , including the following steps:

[0019] Step S301, define the maximum offset value as , reconstruction algorithm parameters The range is , the iterative step length of the reconstruction algorithm is and The storage matrix is ,in, and They are and Pixel offset in direction;

[0020] Step S302: Initialize 、 and reconstruction algorithm parameters ;

[0021] Step S303: Use the reconstruction algorithm parameters The reconstruction algorithm reconstructs the fuzzy coded image and moves the reconstructed fuzzy coded image , will move The blurred coded image after Region is recorded as image ;

[0022] Step S304, calculate the current The image below With image of , and store it in After the storage is completed, ;

[0023] Step S305, repeat steps S303-S304 until completion , then execute step S306, wherein, , and store it in After the storage is completed, ;

[0024] Step S306, reset , and repeat steps S303-S305 until the reconstruction algorithm parameters are completed The pixel offset is , then execute step S307, wherein, , , and store it in After the storage is completed, the reconstruction algorithm parameters are ;

[0025] Step S307, reset , , and repeat steps S303-S306 until the reconstruction algorithm parameters are completed , then execute step S308 and store it in middle;

[0026] Step S308, Get the biggest The index of the largest After the index is converted into the optimal registration parameter, the fuzzy coded image is registered and the fuzzy coded image after registration is recorded as image .

[0027] Furthermore, in the embodiment of the present invention, in step S4, the image Reconstruct the reconstructed image of Save area as image , based on the recorded images With image Temporary structural similarity of , determine the maximum structural similarity The evaluation results of the three-dimensional phase mask encoding imaging system include the following steps:

[0028] Step S401, initialize the maximum structural similarity ;

[0029] Step S402: Use the reconstruction algorithm parameters The reconstruction algorithm for the image Reconstruct the reconstructed image exist Region is recorded as image ;

[0030] Step S403: Record image With image Temporary structural similarity of , if the temporary structural similarity Greater than the maximum structural similarity , then the maximum structural similarity Update to temporary structural similarity Otherwise, the maximum structural similarity is not updated. , let the reconstruction algorithm parameters ,in, is the iterative step size of the reconstruction algorithm;

[0031] Step S404, repeat steps S402-S403 until the reconstruction algorithm parameters are completed After that, the final maximum structural similarity is obtained Evaluation results of the three-dimensional phase mask coded imaging system.

[0032] The image quality evaluation method system of the three-dimensional phase mask imaging system described in the present invention includes the following modules:

[0033] The first acquisition module acquires clear and non-coded images;

[0034] The second acquisition module acquires the fuzzy coded image;

[0035] The registration module registers the clear uncoded image and the fuzzy coded image, and obtains the fuzzy coded image after registration as image ;

[0036] After processing the clear uncoded image, it is recorded as image ;

[0037] Evaluation module, for image Reconstruct the reconstructed image of Save area as image , based on the recorded images With image Temporary structural similarity of , determine the maximum structural similarity Evaluation results of the cubic phase mask coded imaging system.

[0038] An electronic device according to the present invention comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0039] Memory for storing computer programs;

[0040] The processor is configured to implement any of the above-mentioned methods for evaluating image quality of a cubic phase mask imaging system when executing a program stored in the memory.

[0041] The present invention provides a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium. When the computer program is executed by a processor, any of the above-mentioned methods for evaluating the image quality of a cubic phase mask imaging system is implemented.

[0042] The present invention solves the problem that existing evaluation methods lead to low accuracy in image evaluation of a three-dimensional phase mask imaging system. Specific beneficial effects include:

[0043] 1. The image quality evaluation method for a three-dimensional phase mask imaging system described in the present invention addresses the problem of low image evaluation accuracy caused by existing evaluation methods. To address the technical problems of the existing technology, the present invention utilizes the characteristics of sensitivity to pixel misalignment and uses a full-reference evaluation index to evaluate the reconstructed image. Compared with the subjective evaluation or no-reference evaluation index used in the existing technology, the accuracy and robustness are higher.

[0044] 2. The image quality evaluation method of the three-dimensional phase mask imaging system of the present invention combines the reconstruction algorithm of the phase mask coded image and the registration algorithm when registering the reference image and the test image. The indicator is sensitive to pixel misalignment and uses it to correct the misaligned image, solving the problem of incomplete registration of the image of the three-dimensional phase mask coded imaging system caused by the modulation of the three-dimensional phase mask coding plate, providing a prerequisite for the full-reference image quality evaluation of the three-dimensional phase mask coded imaging system. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0046] Figure 1is a diagram of a three-dimensional phase mask coded imaging system according to embodiment 1;

[0047] Figure 2 It is the clear non-coded image described in embodiment 1;

[0048] Figure 3 The fuzzy coded image described in the first embodiment;

[0049] Figure 4 This is the image after registration described in the first embodiment. DETAILED DESCRIPTION

[0050] The following will clearly and completely describe various embodiments of the present invention in conjunction with the accompanying drawings. The embodiments described with reference to the accompanying drawings are exemplary and intended to be used to explain the present invention, but should not be understood as limiting the present invention.

[0051] Embodiment 1: The image quality evaluation method of a three-dimensional phase mask imaging system described in this embodiment includes the following steps:

[0052] Step S1, obtaining a clear and uncoded image;

[0053] Step S2, obtaining a fuzzy coded image;

[0054] Step S3: register the clear uncoded image and the fuzzy coded image, and obtain the registered fuzzy coded image as image ;

[0055] After processing the clear uncoded image, it is recorded as image ;

[0056] Step S4, image Reconstruct the reconstructed image of Save area as image , based on the recorded images With image Temporary structural similarity of , determine the maximum structural similarity Evaluation results of the cubic phase mask coded imaging system.

[0057] In this embodiment, in step S1, obtaining a clear and uncoded image is specifically as follows:

[0058] After the cubic phase mask encoding plate is removed from the cubic phase mask encoding imaging system, the target scene is photographed to obtain a clear non-encoded image.

[0059] In this embodiment, in step S2, the fuzzy coded image is obtained as follows:

[0060] After the cubic phase mask encoding plate is installed in the cubic phase mask encoding imaging system, the target scene is photographed to obtain a fuzzy encoded image.

[0061] In this embodiment, in step S3, the clear uncoded image is processed and recorded as image , specifically:

[0062] Select a clear area within the depth of focus in a clear uncoded image and obtain area, will be clear without coding image Region is recorded as image .

[0063] In this embodiment, in step S3, the clear uncoded image and the fuzzy coded image are registered, and the registered image is recorded as image , including the following steps:

[0064] Step S301, define the maximum offset value as , reconstruction algorithm parameters The range is , the iterative step length of the reconstruction algorithm is and The storage matrix is ,in, and They are and Pixel offset in direction;

[0065] Step S302: Initialize 、 and reconstruction algorithm parameters ;

[0066] Step S303: Use the reconstruction algorithm parameters The reconstruction algorithm reconstructs the fuzzy coded image and moves the reconstructed fuzzy coded image , will move The blurred coded image after Region is recorded as image ;

[0067] Step S304, calculate the current The image below With image of , and store it in After the storage is completed, ;

[0068] Step S305, repeat steps S303-S304 until completion , then execute step S306, wherein, , and store it in After the storage is completed, ;

[0069] Step S306, reset , and repeat steps S303-S305 until the reconstruction algorithm parameters are completed The pixel offset is , then execute step S307, wherein, , , and store it in After the storage is completed, the reconstruction algorithm parameters are ;

[0070] Step S307, reset , , and repeat steps S303-S306 until the reconstruction algorithm parameters are completed , then execute step S308 and store it in middle;

[0071] Step S308, Get the biggest The index of the largest After the index is converted into the optimal registration parameter, the fuzzy coded image is registered and the fuzzy coded image after registration is recorded as image .

[0072] In this embodiment, in the step S4, the image Reconstruct the reconstructed image of Save area as image , based on the recorded images With image Temporary structural similarity of , determine the maximum structural similarity The evaluation results of the three-dimensional phase mask encoding imaging system include the following steps:

[0073] Step S401, initialize the maximum structural similarity ;

[0074] Step S402: Use the reconstruction algorithm parameters The reconstruction algorithm for the image Reconstruct the reconstructed image exist Region is recorded as image ;

[0075] Step S403: Record image With image Temporary structural similarity of , if the temporary structural similarity Greater than the maximum structural similarity , then the maximum structural similarity Updated to temporary structural similarity Otherwise, the maximum structural similarity is not updated. , let the reconstruction algorithm parameters ,in, is the iterative step size of the reconstruction algorithm;

[0076] Step S404, repeat steps S402-S403 until the reconstruction algorithm parameters are completed After that, the final maximum structural similarity is obtained Evaluation results of the three-dimensional phase mask coded imaging system.

[0077] Existing evaluation methods generally use subjective evaluation indicators and no-reference evaluation indicators. However, the image sharpening and ringing effects in the restoration process of the encoded image will lead to low accuracy of image evaluation.

[0078] In order to solve the technical problems existing in the prior art, this embodiment proposes a method for evaluating the image quality of a three-dimensional phase mask imaging system, comprising the following steps:

[0079] The reconstruction algorithm uses the classic Algorithm (spatial domain image restoration algorithm based on Bayesian theory), the reconstruction parameters are ( , the number of iterations), the iteration step is 1, and the optical system parameters are shown in Table 1;

[0080] Table 1 Optical system parameters

[0081]

[0082] Step S1, obtaining a clear and uncoded image, specifically: Figure 1 As shown in Figure 1, the three-dimensional phase mask coded imaging system includes a three-dimensional phase mask encoding plate, an imaging lens, and a detector. The three-dimensional phase mask encoding plate in the three-dimensional phase mask coded optical system is removed and the three-dimensional phase mask coded imaging system is fixed on a stable optical plate to shoot the target scene. The target scene needs to have at least one area within the focal depth of the imaging lens to ensure that a clear area that can be used as a reference image can be obtained. The obtained clear non-coded image is as follows: Figure 2 As shown;

[0083] Step S2, obtaining a fuzzy coded image, specifically: installing the three-dimensional phase mask coding plate back into the three-dimensional phase mask coding imaging system, reshooting the same target scene at the same angle, and obtaining a fuzzy coded image of the same target scene. The obtained fuzzy coded image is as follows: Figure 3 As shown;

[0084] Step S3, registering the clear uncoded image and the blurred coded image, includes the following steps:

[0085] After acquiring a clear uncoded image and a blurred coded image, due to human errors during the installation process and the surface shape of the three-dimensional phase mask coding plate, the clear uncoded image and the blurred coded image cannot meet the image registration requirements of the full reference evaluation. Therefore, before performing the full reference evaluation of the image, it is necessary to combine the imaging principle of the three-dimensional phase mask coding imaging system to align the misaligned coded image and the uncoded image;

[0086] Step S301: Select a clear area within the focus depth in the clear uncoded image and obtain (Rectangular) area , Areas include The horizontal coordinate of the upper left corner of the area, The vertical coordinate of the upper left corner of the area, The width and height of the area will be clear without encoding the image Save area as image ,image is the reference image;

[0087] Step S302, define the maximum offset value as , reconstruction algorithm parameters The range is [1,10], and the iterative step size of the reconstruction algorithm is and The storage matrix is ,in, and They are and Pixel offset in direction;

[0088] Step S303, respectively initialize 、 and reconstruction algorithm parameters ;

[0089] Step S304: Use the reconstruction algorithm parameters The reconstruction algorithm reconstructs the fuzzy coded image and moves the reconstructed fuzzy coded image by the pixel offset , and then move the pixel offset The fuzzy coded image is Region is recorded as image ,image For the test image;

[0090] Step S305: Calculate the current , , The image below With image of ( , an evaluation index in the field of digital image processing) and stored in In order to obtain the best offset later, after the storage is completed, ;

[0091] Step S306, repeat steps S304-S305 until completion , then execute step S307, wherein, , and store it in After the storage is completed, ;

[0092] Step S307, reset , and repeat steps S304-S306 until the reconstruction algorithm parameters are completed The pixel offset is , then execute step S308, wherein, , , and store it in After the storage is completed, the reconstruction algorithm parameters are ;

[0093] Step S308: Reset , , and repeat steps S304-S307 until the reconstruction algorithm parameters are completed , then execute step S309 and store it in middle;

[0094] Step S309, Get the biggest The index of The three-dimensional size of the index is converted into the best registration parameter , After that, the fuzzy coded image is registered and the registered image is recorded as image ;

[0095] Step S4, image Reconstruct the reconstructed image of Save area as image , based on the recorded images With image Temporary structural similarity of , determine the maximum structural similarity The evaluation results of the three-dimensional phase mask encoding imaging system include the following steps:

[0096] Step S401, initialize the maximum structural similarity , reconstruction algorithm parameters ;

[0097] Step S402: Use the reconstruction algorithm parameters The reconstruction algorithm for the image Reconstruct the image and then exist Region is recorded as image ;

[0098] Step S403: Record image With image Temporary structural similarity of , if the temporary structural similarity Greater than the maximum structural similarity , then the maximum structural similarity Updated to temporary structural similarity Otherwise, the maximum structural similarity is not updated. , let the reconstruction algorithm parameters ;

[0099] Step S404, repeat steps S402-S403 until the reconstruction algorithm parameters are completed After that, the final maximum structural similarity is obtained , the final maximum structural similarity As the evaluation result of the three-dimensional phase mask coded imaging system, the reconstructed image at this time is as follows Figure 4 As shown, Figure 4 Compared with the fuzzy coded image without reconstruction registration, Figure 3 As shown, it is obviously closer in the clear area Figure 2 .

[0100] Therefore, the full reference image quality evaluation method of the coded imaging system proposed in this embodiment will obtain the maximum structural similarity As an evaluation result of the cubic phase mask coded imaging system, the maximum structural similarity between the reference image and the test image without the registration method is shown. Significantly more accurate than .

[0101] In summary, the image quality evaluation method of the three-dimensional phase mask imaging system described in the present invention realizes the full-reference image quality evaluation of the three-dimensional phase mask coded imaging system, effectively avoiding the problem of low accuracy of subjective evaluation and no-reference evaluation caused by image sharpening and ringing effects in the restoration process of the coded image. Compared with the currently used no-reference and subjective evaluation methods of the three-dimensional phase mask coded imaging system, the evaluation effect is more accurate and more robust.

[0102] Embodiment 2: The image quality evaluation method system for a three-dimensional phase mask imaging system described in this embodiment includes the following modules:

[0103] The first acquisition module acquires clear and non-coded images;

[0104] The second acquisition module acquires the fuzzy coded image;

[0105] The registration module registers the clear uncoded image and the fuzzy coded image, and obtains the fuzzy coded image after registration as image ;

[0106] After processing the clear uncoded image, it is recorded as image ;

[0107] Evaluation module, for image Reconstruct the reconstructed image of Save area as image , based on the recorded images With image Temporary structural similarity of , determine the maximum structural similarity Evaluation results of the cubic phase mask coded imaging system.

[0108] Embodiment 3: An electronic device described in this embodiment includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0109] Memory for storing computer programs;

[0110] The processor is configured to implement the image quality evaluation method for the cubic phase mask imaging system described in the first embodiment when executing the program stored in the memory.

[0111] Embodiment 4: This embodiment describes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the image quality evaluation method for a cubic phase mask imaging system described in embodiment 1 is implemented.

[0112] The above is a detailed introduction to the image quality evaluation method, system, equipment and medium of the three-dimensional phase mask imaging system proposed in the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for evaluating image quality of a cubic phase mask imaging system, characterized in that: The following steps are involved: Step S1, obtaining a clear and uncoded image; Step S2, obtaining a fuzzy coded image; Step S3, registering the clear uncoded image and the fuzzy coded image, obtaining the registered fuzzy coded image and recording it as image III; After processing the clear and uncoded image, it is recorded as image I; Step S4, reconstructing image III, saving the Rect area of ​​the reconstructed image III as image IIII, and determining the maximum structural similarity best_ssim as the evaluation result of the cubic phase mask coded imaging system based on the recorded temporary structural similarity tmp_ssim between image I and image IIII; In the step S3, the clear uncoded image and the fuzzy coded image are registered to obtain the registered fuzzy coded image, which is recorded as image III, including the following steps: Step S301 , respectively defining the maximum value of the shift as max_shift, the range of the reconstruction algorithm parameter P as [P_min, P_max], the iterative step size of the reconstruction algorithm as Δ, and the PSNR storage matrix as psnr_matrix(x_shift, y_shift, P), where x_shift and y_shift are the pixel shifts in the X and Y directions, respectively; Step S302 , initializing x_shift=-max_shift, y_shift=-max_shift and reconstruction algorithm parameter P=P_min respectively; Step S303, reconstructing the fuzzy coded image using a reconstruction algorithm with reconstruction algorithm parameters P, and shifting the reconstructed fuzzy coded image by (x_shift, y_shift), and recording the fuzzy coded image after shifting by (x_shift, y_shift) as image II in the Rect area; Step S304: Calculate the PSNR of image II and image I under the current x_shift, y_shift, P, and store it in psnr_matrix(x_shift, y_shift, P). After storage, set x_shift = x_shift + 1. Step S305, repeat steps S303-S304 until (x_shift, y_shift) is completed, then execute step S306, wherein x_shift = -max_shift, -max_shift+1, ..., 0, ..., max_shift-1, max_shift, and store it in psnr_matrix (x_shift, y_shift, P). After storage is completed, set y_shift = y_shift+1; Step S306, reset x_shift = -max_shift, and repeat steps S303-S305 until the pixel offset is (x_shift, y_shift) when the reconstruction algorithm parameter P is completed, then execute step S307, wherein x_shift = -max_shift, -max_shift+1, ..., 0, ..., max_shift-1, max_shift, y_shift = -max_shift, -max_shift+1, ..., 0, ..., max_shift-1, max_shift, and store them in psnr_matrix (x_shift, y_shift, P). After the storage is completed, the reconstruction algorithm parameter P is set to P+1; Step S307: reset x_shift = -max_shift, y_shift = -max_shift, and repeat steps S303 to S306 until the algorithm parameters P = P_min, P_min + Δ, ... P_max are reconstructed. Then, step S308 is executed and stored in psnr_matrix (x_shift, y_shift, P). Step S308: Obtain the index of the maximum PSNR in psnr_matrix (x_shift, y_shift, P), convert the index of the maximum PSNR into the optimal registration parameter, and register the fuzzy coded image. The obtained fuzzy coded image after registration is recorded as image III. A clear area within the depth of focus is selected from the clear uncoded image to obtain the Rect area [104, 108, 159, 290]. The Rect area includes the horizontal coordinate of the upper left corner point of the Rect area, the vertical coordinate of the upper left corner point of the Rect area, and the width and height of the Rect area.

2. The image quality evaluation method of a three-dimensional phase mask imaging system according to claim 1, characterized in that: In the step S1, obtaining a clear and uncoded image is specifically as follows: After the cubic phase mask encoding plate is removed from the cubic phase mask encoding imaging system, the target scene is photographed to obtain a clear non-encoded image.

3. The image quality evaluation method of a three-dimensional phase mask imaging system according to claim 1, characterized in that: In the step S2, the fuzzy coded image is obtained, specifically: After the cubic phase mask encoding plate is installed in the cubic phase mask encoding imaging system, the target scene is photographed to obtain a fuzzy encoded image.

4. The image quality evaluation method of a three-dimensional phase mask imaging system according to claim 1, characterized in that: In the step S3, the clear uncoded image is processed and recorded as image I, specifically: A clear area within the focal depth is selected from the clear uncoded image to obtain a Rect area, and the Rect area of ​​the clear uncoded image is recorded as image I.

5. The image quality evaluation method of a three-dimensional phase mask imaging system according to claim 1, characterized in that: In the step S4, the image III is reconstructed, the Rect region of the reconstructed image III is saved as image IIII, and based on the recorded temporary structural similarity tmp_ssim between image I and image IIII, the maximum structural similarity best_ssim is determined as the evaluation result of the cubic phase mask coded imaging system, including the following steps: Step S401, initializing the maximum structural similarity best_ssim=0; Step S402, reconstructing the image III using a reconstruction algorithm with reconstruction algorithm parameters P, and recording the reconstructed image III as image IIII in the Rect area; Step S403: record the temporary structural similarity tmp_ssim between image I and image IIII. If the temporary structural similarity tmp_ssim is greater than the maximum structural similarity best_ssim, update the maximum structural similarity best_ssim to the temporary structural similarity tmp_ssim. Otherwise, do not update the maximum structural similarity best_ssim. Set the reconstruction algorithm parameter P = P + Δ, where Δ is the iterative step size of the reconstruction algorithm. Step S404 , repeating steps S402 to S403 until the reconstruction algorithm parameters P=P_min, P_min+Δ, …P_max are completed, and obtaining the final maximum structural similarity best_ssim as the evaluation result of the cubic phase mask coded imaging system.

6. A method and system for evaluating image quality of a cubic phase mask imaging system, characterized in that: Includes the following modules: The first acquisition module acquires clear and non-coded images; The second acquisition module acquires the fuzzy coded image; The registration module registers the clear uncoded image and the fuzzy coded image, and obtains the registered fuzzy coded image, which is recorded as image III; After processing the clear and uncoded image, it is recorded as image I; The evaluation module reconstructs the image III and saves the Rect region of the reconstructed image III as image IIII. Based on the recorded temporary structural similarity tmp_ssim between image I and image IIII, the maximum structural similarity best_ssim is determined as the evaluation result of the cubic phase mask coded imaging system. In the registration module, the clear uncoded image and the fuzzy coded image are registered, and the fuzzy coded image obtained after registration is recorded as image III, which includes the following modules: Module S301 defines the maximum value of the shift as max_shift, the range of the reconstruction algorithm parameter P as [P_min, P_max], the iterative step size of the reconstruction algorithm as Δ, and the PSNR storage matrix as psnr_matrix(x_shift, y_shift, P), where x_shift and y_shift are the pixel shifts in the X and Y directions respectively. Module S302, respectively initialize x_shift = -max_shift, y_shift = -max_shift and reconstruction algorithm parameter P = P_min; Module S303, reconstructing the fuzzy coded image using a reconstruction algorithm with reconstruction algorithm parameters P, and shifting the reconstructed fuzzy coded image by (x_shift, y_shift), and recording the fuzzy coded image after shifting by (x_shift, y_shift) as image II in the Rect area; Module S304 calculates the PSNR of image II and image I under the current x_shift, y_shift, P, and stores it in psnr_matrix(x_shift, y_shift, P). After storage, set x_shift = x_shift + 1. Module S305 repeats modules S303-S304 until (x_shift, y_shift) is completed, then executes step S306, wherein x_shift = -max_shift, -max_shift+1, ..., 0, ..., max_shift-1, max_shift, and stores it in psnr_matrix (x_shift, y_shift, P). After storage is completed, set y_shift = y_shift+1; Module S306 resets x_shift = -max_shift and repeats modules S303 to S305 until the pixel offset is (x_shift, y_shift) when the algorithm parameter P is reconstructed. Module S307 is then executed, wherein x_shift = -max_shift, -max_shift+1, ..., 0, ..., max_shift-1, max_shift, y_shift = -max_shift, -max_shift+1, ..., 0, ..., max_shift-1, max_shift, and the pixels are stored in psnr_matrix (x_shift, y_shift, P). After the storage is completed, the reconstruction algorithm parameter P is set to P+1. Module S307, reset x_shift = -max_shift, y_shift = -max_shift, and repeat steps S303 to S306 until the reconstruction algorithm parameters P = P_min, P_min + Δ, ... P_max are completed, then execute module S308 and store them in psnr_matrix (x_shift, y_shift, P); Module S308, obtaining the index of the maximum PSNR in psnr_matrix (x_shift, y_shift, P), converting the index of the maximum PSNR into an optimal registration parameter, registering the fuzzy coded image, and obtaining the registered fuzzy coded image, recording it as image III; A clear area within the depth of focus is selected from the clear uncoded image to obtain the Rect area [104, 108, 159, 290]. The Rect area includes the horizontal coordinate of the upper left corner point of the Rect area, the vertical coordinate of the upper left corner point of the Rect area, and the width and height of the Rect area.

7. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory for storing computer programs; The processor is configured to implement the image quality evaluation method for a three-dimensional phase mask imaging system according to any one of claims 1 to 5 when executing a program stored in the memory.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the image quality evaluation method for a cubic phase mask imaging system according to any one of claims 1 to 5 is implemented.

Images