Intelligent hologram compression and reconstruction system and method

Abstract

The invention relates to a hologram intelligent compression and reconstruction system and method, and belongs to the field of hologram processing. Comprising the step of realizing high-efficiency compression and high-quality reconstruction of holographic data through three core technologies of dynamic range segmentation, perception weight optimization and adaptive quantization when a CCD records a hologram. The method comprises the following steps: firstly, segmenting a hologram into a high dynamic range (HDR) region and a low dynamic range (LDR) region by adopting a local feature analysis technology and combining a dynamic threshold value; secondly, enhancing high-frequency information retention by using a frequency sensitive weight; and iteratively optimizing parameters through a code rate control module, implementing nonlinear quantization on an HDR region, and implementing linear quantization on an LDR region. In the reconstruction stage, a Fresnel diffraction model is adopted for physical reduction, and a holographic reproductive image is obtained. According to the method, a high compression ratio is achieved in airspace compression, and meanwhile high quality of a reconstructed image of the compressed hologram is guaranteed.

Description

Technical Field

[0001] The present invention relates to a hologram intelligent compression and reconstruction system and method, which are used for compressing and reconstructing holographic three-dimensional display data and belong to the technical field of hologram processing. Background Art

[0002] Holographic technology, by fully recording the amplitude and phase information of an object's light wavefield, can accurately reconstruct the wavefront characteristics of a three-dimensional scene. It has been widely used in key areas such as holographic microscopy and true three-dimensional display. In recent years, three-dimensional display technology has rapidly advanced. Holographically reconstructed three-dimensional scenes can restore the true light field distribution and provide complete depth information, fundamentally overcoming the limitations of binocular parallax. It is recognized as the optimal technical route for achieving true stereoscopic display. With the evolution of 5G / 6G communications and virtual reality technologies, real-time three-dimensional holographic display is becoming a reality. However, high-resolution, complete three-dimensional image reconstruction requires the acquisition of multi-angle holograms of the object. The resulting massive amount of data severely restricts real-time transmission and display efficiency. Implementing holographic data compression has become a key approach to overcoming this bottleneck, significantly reducing the computational load and improving transmission efficiency.

[0003] Traditional digital image coding standards rely on models of human vision, but holograms are composed of complex interference fringes. Their data structure differs significantly from natural images, containing numerous high-frequency components that significantly impact reconstruction quality, rendering conventional coding standards ineffective. While various holographic compression schemes have been proposed by researchers both domestically and internationally, four major drawbacks persist: high algorithmic complexity, lengthy processing times, insufficient compression ratios, and degraded reconstruction quality. Consequently, a unified technical standard for efficient holographic compression has yet to be established. Summary of the Invention

[0004] The present invention provides a hologram intelligent compression and reconstruction system and method, which is used to solve the problems of current hologram compression algorithms being complex, time-consuming, insufficient compression ratio, and poor hologram reconstruction quality. While the present invention has a high compression ratio, it also ensures the high quality of the compressed hologram reconstructed image.

[0005] The technical solution of the present invention is as follows: On one hand, the present invention provides a hologram intelligent compression and reconstruction system, which includes the following core processing units working in coordination:

[0006] Dynamic range segmentation module: When the CCD records the hologram, it uses local feature analysis technology with a configurable window size. By calculating the statistical characteristics of the local area of ​​the hologram, it generates a dynamic range segmentation threshold parameter, dividing the image into a high dynamic range (HDR) area containing complex details and a low dynamic range (LDR) area with a smooth background. The dynamic range segmentation threshold parameter is optimized in real time through a closed-loop feedback mechanism.

[0007] Visual Perception Compression Module: This module includes a frequency domain enhancement unit and a partition quantization unit. The frequency domain enhancement unit selectively enhances spatial frequency information based on the human visual sensitivity model. The partition quantization unit uses high-bit-depth nonlinear transformation compression for HDR areas and low-bit-depth linear compression for LDR areas to form a mixed bit-depth representation.

[0008] Closed-loop bitrate control module: dynamically adjusts the dynamic range segmentation threshold parameters through multiple rounds of iterative optimization algorithms;

[0009] Physical reconstruction and interactive evaluation module: Based on the Fresnel-Kirchhoff diffraction integral model, the compressed data is restored to a visual holographic image, and a comprehensive evaluation report including objective quality indicators and compression efficiency is simultaneously generated, and a real-time adjustable interactive visualization interface is provided.

[0010] Furthermore, the dynamic range segmentation module specifically performs the following functions:

[0011] Define a sliding analysis window (size adjustable from 3*3 to 7*7 pixels) and calculate the statistical characteristics of the local area of ​​the hologram region by region;

[0012] According to the statistical distribution characteristics of the entire image, combined with the preset weight coefficient, an adaptive dynamic range segmentation threshold parameter is generated;

[0013] Output binary mask data, clearly identifying the spatial distribution of HDR areas (value 1) and LDR areas (value 0).

[0014] Furthermore, the visual perception compression module includes:

[0015] Frequency domain enhancement unit: used to construct a spatial frequency response function model, enhance the mid- and high-frequency information that the human eye is sensitive to, and suppress high-frequency noise that the visual sense is not sensitive to;

[0016] Adaptive quantization unit: used to implement 8-12 bit logarithmic domain nonlinear quantization for HDR areas to retain high-intensity gradient details; implement 2-4 bit linear quantization for LDR areas to efficiently compress smooth background areas.

[0017] Furthermore, the closed-loop bit rate control module includes:

[0018] Parameter space configuration unit: used to initialize the feasible search range of dynamic range segmentation threshold parameters, where the main coefficient α∈[2.0,3.0] and the offset coefficient β∈[0.1,0.3];

[0019] Iterative convergence unit: used to gradually shrink the parameter space using a binary search strategy, and calculate the actual compression bit rate of the current parameter combination in each round of iteration;

[0020] Dynamic adjustment unit: used to dynamically adjust the parameter search boundary according to the deviation direction between the actual bit rate and the target bit rate;

[0021] Optimization output unit: used to output the optimal parameter combination that makes the bit rate error ≤ 0.02bpp within a preset maximum number of iterations (10-20 times).

[0022] Furthermore, the physical reconstruction and interactive evaluation module includes:

[0023] Optical reconstruction unit: used to calculate the light wave propagation process through the Fresnel-Kirchhoff diffraction integral model and realize holographic image reconstruction in combination with fast Fourier transform;

[0024] Interactive control unit: used to provide a graphical interface for users to adjust the contrast of reconstructed images in real time and dynamically optimize visual effects;

[0025] Quality assessment unit: used to simultaneously calculate the three quality indicators of peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and perceptual similarity (LPIPS), and calculate the compression ratio based on the original bit depth and compressed bit depth.

[0026] Furthermore, the quality assessment unit includes a data volume calculation subunit, a compression efficiency subunit, and a quality verification subunit;

[0027] In the data amount calculation subunit, original data amount = total number of pixels * original bit depth (8-16 bits);

[0028] Compressed data volume = HDR pixel count * HDR quantization bit count + LDR pixel count * LDR quantization bit count;

[0029] In the compression efficiency subunit, compression ratio = original data volume / compressed data volume ≥ 8.0:1;

[0030] In the quality verification subunit, it is ensured that the reconstructed image meets the quality standards of PSNR>25dB and SSIM>0.9.

[0031] Furthermore, it also includes a pre-processing module:

[0032] Data parsing unit: used to support hologram input in TIFF / RAW / BMP formats and automatically identify the original bit depth and image size;

[0033] Physical parameter configuration unit: used to configure the core parameters of optical reconstruction: laser wavelength (500-550nm), pixel size (0.003-0.006mm), reconstruction distance (1000-2000mm);

[0034] Calibration unit: used to perform grayscale normalization on the input image to eliminate the response differences of the acquisition equipment.

[0035] In a second aspect, the present invention further provides a method for intelligent compression and reconstruction of a hologram, the method comprising:

[0036] Data preparation and dynamic segmentation: When recording holograms on a CCD, a local feature analysis technique with a configurable window size is used. The dynamic range segmentation threshold parameter is generated by calculating the statistical characteristics of the local area of ​​the hologram, dividing the image into HDR and LDR areas. The dynamic range segmentation threshold parameter is optimized in real time through a closed-loop feedback mechanism.

[0037] Visual perception compression: includes a frequency domain enhancement unit and a partition quantization unit; the frequency domain enhancement unit selectively enhances spatial frequency information based on the human eye visual sensitivity model; the partition quantization unit uses high-bit-depth nonlinear transformation compression for HDR areas and low-bit-depth linear compression for LDR areas to form a mixed bit-depth representation;

[0038] Bitrate optimization: Dynamically adjust the dynamic range segmentation threshold parameters through multiple rounds of iterative optimization algorithms;

[0039] Physical reconstruction and interactive evaluation module: Based on the Fresnel-Kirchhoff diffraction integral model, the compressed data is restored to a visual holographic image, and a comprehensive evaluation report including objective quality indicators and compression efficiency is simultaneously generated, and a real-time adjustable interactive visualization interface is provided.

[0040] Furthermore, the bit rate optimization includes the following steps:

[0041] (1) Initialize the dynamic range segmentation threshold parameter search space: main coefficient α∈[α_min,α_max], offset coefficient β∈[β_min,β_max]; where α_min and α_max represent the minimum and maximum values ​​of the main coefficient respectively; β_min and β_max represent the minimum and maximum values ​​of the offset coefficient respectively;

[0042] (2) Calculate the midpoint value of the current dynamic range segmentation threshold parameter:

[0043] α_mid=(α_min+α_max) / 2, β_mid=(β_min+β_max) / 2;

[0044] Among them, α_mid and β_mid represent the midpoint values ​​of the main coefficient and offset coefficient of the dynamic range segmentation threshold parameter respectively;

[0045] (3) Calculate the actual compression bit rate based on the current dynamic range segmentation threshold parameter;

[0046] If the actual bitrate > target bitrate + tolerance threshold, update:

[0047] α_min=α_mid, β_min=β_mid;

[0048] If the actual bitrate is less than the target bitrate - tolerance threshold, update:

[0049] α_max=α_mid,β_max=β_mid;

[0050] (4) Repeat (2)-(3) until the bit rate accuracy requirement is met or the maximum number of iterations is reached; output the final parameter combination (α_opt, β_opt).

[0051] In a third aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which implements the method for intelligent compression and reconstruction of holograms when executed by a processor.

[0052] The beneficial effects of the present invention are:

[0053] 1. The present invention proposes a hologram intelligent compression and reconstruction system, which can directly compress and output the image after recording the hologram;

[0054] 2. The compression technology of this invention is universally applicable to both real-time acquired optical holograms and computer-generated digital holograms, achieving a significant improvement in compression ratio while ensuring high-quality reconstruction. Its high compatibility and superior performance provide an innovative technical path for coding standardization in fields such as holographic communications and three-dimensional displays.

[0055] 3. The present invention proposes an intelligent hologram compression and reconstruction system, which effectively realizes the compression of holograms. This compression method has a high compression ratio while also ensuring the high quality of the reconstructed hologram image after compression.

[0056] 4. The complete hologram recording, compression, and reconstruction system proposed in the present invention, namely, a hologram compression system based on dynamic range adaptive quantization and perceptual joint coding, provides a new approach for the realization of future hologram coding standards; the present invention reduces the communication bandwidth pressure for real-time transmission and display of three-dimensional holographic images. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 Schematic diagram of the optical path for collecting the hologram;

[0058] Figure 2 is the collected original color hologram;

[0059] Figure 3 A flowchart of the compression method proposed by the present invention;

[0060] Figure 4 Flow chart of the whole system of the present application;

[0061] Figure 5 Comparison diagram of the compression method of the present application and wavelet transform.

[0062] Legend: BS1, BS2 - beam splitter, MO1, MO2 - objective lens, M1, M2, M3 - mirror, Laser R, Laser G, Laser B - red, green, blue laser. DETAILED DESCRIPTION

[0063] Example 1: As shown, the present application provides a hologram intelligent compression and reconstruction system, and the present application provides a hologram intelligent compression and reconstruction system, which comprises the following core processing units working cooperatively: Figure 1-Figure 5 Dynamic range segmentation module: local feature analysis technology with configurable window size is adopted when recording hologram by CCD, dynamic range segmentation threshold parameters are generated by calculating statistical characteristics of local area of hologram, and the image is divided into HDR area and LDR area, wherein the dynamic range segmentation threshold parameters are optimized in real time through closed loop feedback mechanism;

[0064] Visual perception compression module: including frequency domain enhancement unit and partition quantization unit; the frequency domain enhancement unit selectively strengthens spatial frequency information based on human eye visual sensitivity model; the partition quantization unit adopts high bit depth nonlinear transformation compression for HDR area and low bit depth linear compression for LDR area, forming a mixed bit depth representation;

[0065] Closed loop code rate control module: dynamically adjusts the dynamic range segmentation threshold parameters through multi-round iteration optimization algorithm;

[0066] Physical reconstruction and interactive evaluation module: based on Fresnel-Kirchhoff diffraction integral model, the compressed data is restored to a visible holographic image, a comprehensive evaluation report containing objective quality indicators and compression efficiency is generated synchronously, and a real-time adjustable interactive visualization interface is provided.

[0067] Further, the dynamic range segmentation module specifically performs the following functions:

[0068] Define a sliding analysis window, and calculate the statistical characteristics of the local area of the hologram region by region;

[0069] According to the statistical characteristic distribution characteristics of the whole image, combined with the preset weight coefficient, the adaptive dynamic range segmentation threshold parameters are generated;

[0070] Output binary mask data to clearly identify the spatial distribution of HDR area and LDR area.

[0071] ​

[0072] Furthermore, the visual perception compression module includes:

[0073] Frequency domain enhancement unit: used to construct a spatial frequency response function model, enhance the mid- and high-frequency information that the human eye is sensitive to, and suppress high-frequency noise that the visual sense is not sensitive to;

[0074] Adaptive quantization unit: used to implement 8-12 bit logarithmic domain nonlinear quantization for HDR areas to retain high-intensity gradient details; implement 2-4 bit linear quantization for LDR areas to efficiently compress smooth background areas.

[0075] Furthermore, the closed-loop bit rate control module includes:

[0076] Parameter space configuration unit: used to initialize the feasible search range of the dynamic range segmentation threshold parameter, where the main coefficient α∈[2.0,3.0] and the offset coefficient α∈[0.1,0.3];

[0077] Iterative convergence unit: used to gradually shrink the parameter space using a binary search strategy, and calculate the actual compression bit rate of the current parameter combination in each round of iteration;

[0078] Dynamic adjustment unit: used to dynamically adjust the parameter search boundary according to the deviation direction between the actual bit rate and the target bit rate;

[0079] Optimization output unit: used to output the optimal parameter combination that makes the bit rate error ≤ 0.02bpp within a preset maximum number of iterations.

[0080] Furthermore, the physical reconstruction and interactive evaluation module includes:

[0081] Optical reconstruction unit: used to calculate the light wave propagation process through the Fresnel-Kirchhoff diffraction integral model and realize holographic image reconstruction in combination with fast Fourier transform;

[0082] Interactive control unit: used to provide a graphical interface for users to adjust the contrast of reconstructed images in real time and dynamically optimize visual effects;

[0083] Quality assessment unit: used to simultaneously calculate the three quality indicators of peak signal-to-noise ratio, structural similarity, and perceptual similarity, and calculate the compression ratio based on the original bit depth and compressed bit depth.

[0084] Furthermore, the quality assessment unit includes a data volume calculation subunit, a compression efficiency subunit, and a quality verification subunit;

[0085] In the data amount calculation subunit, original data amount = total number of pixels * original bit depth;

[0086] Compressed data volume = HDR pixel count * HDR quantization bit count + LDR pixel count * LDR quantization bit count;

[0087] In the compression efficiency subunit, the compression ratio = original data amount / compressed data amount >= 8.0:1;

[0088] In the quality verification subunit, the reconstructed image satisfies the quality standards of PSNR > 25 dB and SSIM > 0.9.

[0089] Further, the pre-processing module is further included.

[0090] The data analysis unit is used for supporting the input of holograms in TIFF / RAW / BMP formats, automatically identifying the original bit depth and image size;

[0091] The physical parameter configuration unit is used for configuring the optical reconstruction core parameters, including the laser wavelength (500-550 nm), the pixel size (0.003-0.006 mm) and the reconstruction distance (1000-2000 mm).

[0092] The calibration unit is used for carrying out the gray scale normalization processing on the input image, and eliminating the response difference of the acquisition device.

[0093] In the second aspect, the application further provides a hologram intelligent compression and reconstruction method, and the method comprises the following steps:

[0094] Data preparation and dynamic segmentation: when the CCD records the hologram, the local feature analysis technology with a configurable window size is adopted, the dynamic range segmentation threshold parameter is generated by calculating the statistical characteristics of the local area of the hologram, the image is divided into the HDR area and the LDR area, and the dynamic range segmentation threshold parameter is optimized in real time through the closed loop feedback mechanism.

[0095] During the data preparation, first, the experimental optical path device required for the hologram acquisition is built, as shown in Figure 1 The used laser has a red light wavelength of 671 nm, a green light wavelength of 532 nm and a blue light wavelength of 457 nm, the CCD has a resolution of 1344 pixel x 1024 pixel, the model is TriCCD_Hamamatsu_ONERA C7780-20, the pixel size is 6.45 um x 6.45 um, the distance between the object and the CCD is 2 m, and the obtained hologram is as shown in Figure 2 Then, the obtained hologram is compressed.

[0096] Visual perception compression: the visual perception compression comprises a frequency domain enhancement unit and a partition quantization unit; the frequency domain enhancement unit selectively strengthens the spatial frequency information based on the human eye visual sensitivity model; the partition quantization unit adopts the high bit depth nonlinear transformation compression for the HDR area and adopts the low bit depth linear compression for the LDR area, so as to form the mixed bit depth representation.

[0097] Bitrate optimization: Dynamically adjust the dynamic range segmentation threshold parameters through multiple rounds of iterative optimization algorithms;

[0098] Physical reconstruction and interactive evaluation module: Based on the Fresnel-Kirchhoff diffraction integral model, the compressed data is restored to a visual holographic image, and a comprehensive evaluation report including objective quality indicators and compression efficiency is simultaneously generated, and a real-time adjustable interactive visualization interface is provided.

[0099] Furthermore, the bit rate optimization includes the following steps:

[0100] (1) Initialize the dynamic range segmentation threshold parameter search space: main coefficient α∈[α_min,α_max], offset coefficient β∈[β_min,β_max]; where α_min and α_max represent the minimum and maximum values ​​of the main coefficient respectively; β_min and β_max represent the minimum and maximum values ​​of the offset coefficient respectively;

[0101] (2) Calculate the midpoint value of the current dynamic range segmentation threshold parameter:

[0102] α_mid=(α_min+α_max) / 2, β_mid=(β_min+β_max) / 2;

[0103] Among them, α_mid and β_mid represent the midpoint values ​​of the main coefficient and offset coefficient of the dynamic range segmentation threshold parameter respectively;

[0104] (3) Calculate the actual compression bit rate based on the current dynamic range segmentation threshold parameter;

[0105] If the actual bitrate > target bitrate + tolerance threshold, update:

[0106] α_min=α_mid, β_min=β_mid;

[0107] If the actual bitrate is less than the target bitrate - tolerance threshold, update:

[0108] α_max=α_mid,β_max=β_mid;

[0109] (4) Repeat (2)-(3) until the bit rate accuracy requirement is met or the maximum number of iterations is reached; output the final parameter combination (α_opt, β_opt).

[0110] In a third aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which implements the method for intelligent compression and reconstruction of holograms when executed by a processor.

[0111] In order to objectively evaluate the reconstruction quality of the compressed hologram, the present invention introduces the following key indicators:

[0112] Peak Signal-to-Noise Ratio (PSNR) is an objective evaluation metric for image reconstruction quality, quantifying the ratio between the peak signal and the distortion noise. A higher PSNR value indicates better reconstructed image quality.

[0113] The Multi-Scale Structural Similarity Index Measure (MS-SSIM), a quality assessment method that more closely reflects human subjective perception, assesses image similarity through multi-scale analysis. It iteratively applies low-pass filtering to the reference image and the compressed image, then performs structural comparisons at different scales after progressive downsampling. The resulting score is a composite score in the range [0, 1], with higher values ​​indicating closer reconstruction quality to the original image.

[0114] Compression ratio (CR), the core indicator for measuring compression efficiency, is defined as the ratio of the original image data volume to the compressed image data volume.

[0115] Table 1 shows the compression effect of the present invention on the actual photographed hologram

[0116]

[0117] Figure 5 The comparison between the compressed hologram and the original hologram reconstructed image is shown. The left side is the reconstructed image of the original image, the middle is the reconstructed image of the hologram after compression by the method of the present invention, and the rightmost side is the reconstructed image of the hologram after compression by wavelet transform.

[0118] It can be found that the PSNR of the hologram compressed by the method of the present invention can still reach 27.66 when the compression ratio reaches 8.0, which is a significant improvement compared to directly using wavelet transform. While the compression ratio is relatively large, the reconstruction quality of the hologram is guaranteed.

[0119] This invention proposes a system and method for intelligent hologram compression and reconstruction, enabling direct image compression and output after hologram recording. This compression technology is universally applicable to both real-time optical holograms and computer-generated digital holograms, significantly improving compression ratios while ensuring high-quality reconstruction. Its high compatibility and superior performance provide an innovative technical path for coding standardization in areas such as holographic communications and three-dimensional displays.

[0120] The specific embodiments of the present invention are described in detail above with reference to the accompanying drawings. However, 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.

Claims

1. A hologram intelligent compression and reconstruction system, characterized in that: The system includes the following core processing units working together: Dynamic range segmentation module: When the CCD records the hologram, it uses local feature analysis technology with a configurable window size. By calculating the statistical characteristics of the local area of ​​the hologram, it generates a dynamic range segmentation threshold parameter, dividing the image into HDR and LDR areas. The dynamic range segmentation threshold parameter is optimized in real time through a closed-loop feedback mechanism. Visual perception compression module: includes frequency domain enhancement unit and partition quantization unit; The frequency domain enhancement unit selectively enhances spatial frequency information based on the human visual sensitivity model; The partition quantization unit uses high-bit-depth nonlinear transformation compression for HDR areas and low-bit-depth linear compression for LDR areas to form a mixed bit-depth representation; Closed-loop bitrate control module: dynamically adjusts the dynamic range segmentation threshold parameters through multiple rounds of iterative optimization algorithms; Physical reconstruction and interactive evaluation module: Based on the Fresnel-Kirchhoff diffraction integral model, the compressed data is restored to a visual holographic image, and a comprehensive evaluation report including objective quality indicators and compression efficiency is simultaneously generated, and a real-time adjustable interactive visualization interface is provided.

2. The hologram intelligent compression and reconstruction system according to claim 1, wherein: The dynamic range segmentation module specifically performs the following functions: Define a sliding analysis window and calculate the statistical characteristics of the local area of ​​the hologram region by region; According to the statistical distribution characteristics of the entire image, combined with the preset weight coefficient, an adaptive dynamic range segmentation threshold parameter is generated; Output binary mask data to clearly identify the spatial distribution of HDR and LDR areas.

3. The hologram intelligent compression and reconstruction system according to claim 1, wherein: The visual perception compression module includes: Frequency domain enhancement unit: used to construct a spatial frequency response function model, enhance the mid- and high-frequency information that the human eye is sensitive to, and suppress high-frequency noise that the visual sense is not sensitive to; Adaptive quantization unit: used to implement 8-12 bit logarithmic domain nonlinear quantization for HDR areas to retain high-intensity gradient details; implement 2-4 bit linear quantization for LDR areas to efficiently compress smooth background areas.

4. The hologram intelligent compression and reconstruction system according to claim 1, wherein: The closed-loop bit rate control module includes: Parameter space configuration unit: used to initialize the feasible search range of dynamic range segmentation threshold parameters, where the main coefficient α∈[2.0,3.0] and the offset coefficient β∈[0.1,0.3]; Iterative convergence unit: used to gradually shrink the parameter space using a binary search strategy, and calculate the actual compression bit rate of the current parameter combination in each round of iteration; Dynamic adjustment unit: used to dynamically adjust the parameter search boundary according to the deviation direction between the actual bit rate and the target bit rate; Optimization output unit: used to output the optimal parameter combination that makes the bit rate error ≤ 0.02bpp within a preset maximum number of iterations.

5. The hologram intelligent compression and reconstruction system according to claim 1, wherein: The physical reconstruction and interactive evaluation module includes: Optical reconstruction unit: used to calculate the light wave propagation process through the Fresnel-Kirchhoff diffraction integral model and realize holographic image reconstruction in combination with fast Fourier transform; Interactive control unit: used to provide a graphical interface for users to adjust the contrast of reconstructed images in real time and dynamically optimize visual effects; Quality assessment unit: used to simultaneously calculate the three quality indicators of peak signal-to-noise ratio, structural similarity, and perceptual similarity, and calculate the compression ratio based on the original bit depth and compressed bit depth.

6. The hologram intelligent compression and reconstruction system according to claim 5, characterized in that: The quality assessment unit includes a data volume calculation subunit, a compression efficiency subunit, and a quality verification subunit; In the data amount calculation subunit, original data amount = total number of pixels * original bit depth; Compressed data volume = HDR pixel count * HDR quantization bit count + LDR pixel count * LDR quantization bit count; In the compression efficiency subunit, compression ratio = original data volume / compressed data volume ≥ 8.0:1; In the quality verification subunit, it is ensured that the reconstructed image meets the quality standards of PSNR>25dB and SSIM>0.

9.

7. The hologram intelligent compression and reconstruction system according to claim 1, wherein: Also includes preprocessing modules: Data parsing unit: used to support hologram input in TIFF / RAW / BMP formats and automatically identify the original bit depth and image size; Physical parameter configuration unit: used to configure the core parameters of optical reconstruction: laser wavelength (500-550nm), pixel size (0.003-0.006mm), reconstruction distance (1000-2000mm); Calibration unit: used to perform grayscale normalization on the input image to eliminate the response differences of the acquisition equipment.

8. A hologram intelligent compression and reconstruction method, characterized in that: The method comprises: Data preparation and dynamic segmentation: When recording holograms on a CCD, a local feature analysis technique with a configurable window size is used. The dynamic range segmentation threshold parameter is generated by calculating the statistical characteristics of the local area of ​​the hologram, dividing the image into HDR and LDR areas. The dynamic range segmentation threshold parameter is optimized in real time through a closed-loop feedback mechanism. Visual perception compression: includes a frequency domain enhancement unit and a partition quantization unit; the frequency domain enhancement unit selectively enhances spatial frequency information based on the human eye visual sensitivity model; the partition quantization unit uses high-bit-depth nonlinear transformation compression for HDR areas and low-bit-depth linear compression for LDR areas to form a mixed bit-depth representation; Bitrate optimization: Dynamically adjust the dynamic range segmentation threshold parameters through multiple rounds of iterative optimization algorithms; Physical reconstruction and interactive evaluation module: Based on the Fresnel-Kirchhoff diffraction integral model, the compressed data is restored to a visual holographic image, and a comprehensive evaluation report including objective quality indicators and compression efficiency is simultaneously generated, and a real-time adjustable interactive visualization interface is provided.

9. The method for intelligent compression and reconstruction of holograms according to claim 8, characterized in that: The bit rate optimization comprises the following steps: (1) Initialize the dynamic range segmentation threshold parameter search space: main coefficient α∈[α_min,α_max], offset coefficient β∈[β_min,β_max]; where α_min and α_max represent the minimum and maximum values ​​of the main coefficient respectively; β_min and β_max represent the minimum and maximum values ​​of the offset coefficient respectively; (2) Calculate the midpoint value of the current dynamic range segmentation threshold parameter: α_mid=(α_min+α_max) / 2, β_mid=(β_min+β_max) / 2; Among them, α_mid and β_mid represent the midpoint values ​​of the main coefficient and offset coefficient of the dynamic range segmentation threshold parameter respectively; (3) Calculate the actual compression bit rate based on the current dynamic range segmentation threshold parameter; If the actual bitrate > target bitrate + tolerance threshold, update: α_min=α_mid, β_min=β_mid; If the actual bitrate is less than the target bitrate - tolerance threshold, update: α_max=α_mid,β_max=β_mid; (4) Repeat (2)-(3) until the bit rate accuracy requirement is met or the maximum number of iterations is reached; output the final parameter combination (α_opt, β_opt).

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, a hologram intelligent compression and reconstruction method as described in any one of claims 8 to 9 is implemented.

Images