Naked eye 3D screen and 3D image display method

Abstract

The invention relates to the technical field of 3D image display, in particular to a naked eye 3D screen and a 3D image display method. The method comprises the following steps: carrying out background-foreground main body 3D space geometric model construction on a multimedia video playing image to obtain a background-foreground main body 3D space geometric model; performing visual angle movement light change illusion quantization on the background-foreground main body 3D space geometric model to obtain visual angle movement light change illusion quantization data; performing parallax focusing gradient loss evaluation according to the visual angle movement light change illusion quantization data to obtain parallax focusing gradient loss data; designing an optical variable regulation and control mechanism according to the parallax focusing gradient loss data to obtain the optical variable regulation and control mechanism; performing logic memory learning on the light variable regulation and control mechanism based on a strategy gradient algorithm to obtain light variable regulation and control logic memory data; according to the invention, the 3D image display technology is optimized, so that the 3D image display technology is more perfect.

Description

Technical Field

[0001] The present invention relates to the technical field of 3D image display, and in particular to a naked-eye 3D screen and a 3D image display method. Background Art

[0002] 3D image display technology is the intersection of information technology, visual science and graphics. It aims to convert two-dimensional image data into display content with spatial depth and three-dimensionality by simulating the human eye's visual perception of the real world. The background of this technology comes from the pursuit of a more realistic reproduction of the natural world. From the invention of perspective in artistic painting to the development of modern computer-generated virtual reality (VR) and augmented reality (AR), it reflects the emphasis on three-dimensional spatial expression. 3D image display methods not only occupy an important position in the entertainment field (such as movies, games and animation), but are also widely used in medical imaging, industrial design, aerospace, military simulation and other fields, becoming an important driving force for the development of science and technology and society. However, a traditional 3D image display method has the problems of low accuracy in visual fatigue analysis and large error in regulation accuracy. Summary of the invention

[0003] Based on this, it is necessary to provide a naked-eye 3D screen and a 3D image display method to solve at least one of the above technical problems.

[0004] To achieve the above object, a 3D image display method is provided, the method comprising the following steps:

[0005] Step S1: Acquire a multimedia video playback image; construct a background-foreground subject 3D space geometry model for the multimedia video playback image to obtain a background-foreground subject 3D space geometry model;

[0006] Step S2: obtaining user perspective movement data; performing perspective movement light change illusion quantification on the background-foreground subject 3D space geometric model according to the user perspective movement data to obtain perspective movement light change illusion quantification data; performing parallax focus gradient loss evaluation according to the perspective movement light change illusion quantification data to obtain parallax focus gradient loss data;

[0007] Step S3: evaluating visual perception fatigue according to the parallax focus gradient loss data to obtain visual perception fatigue data; designing a light variable control mechanism for the perspective movement light change illusion quantification data and the parallax focus gradient loss data based on the visual perception fatigue data to obtain a light variable control mechanism;

[0008] Step S4: Based on the policy gradient algorithm, logic memory learning is performed on the optically variable control mechanism to obtain optically variable control logic memory data; according to the optically variable control logic memory data, automated control firmware design is performed to obtain optically variable control logic firmware, and the optically variable control logic firmware is embedded into the naked-eye 3D screen image control processor to execute the 3D image display method.

[0009] The present invention obtains the image of multimedia video playback, uses computer vision and deep learning algorithms to extract the spatial information of the background and foreground subjects (i.e., the theme elements in the image, the protagonist, etc.), and performs three-dimensional geometric modeling. This process can convert the image content in the video into a 3D data structure that can be deeply analyzed and processed, thereby providing a basic geometric framework for subsequent operations such as viewing angle adjustment and light changes, so that the subsequent visual effects can be more accurate and natural. The user's viewing angle movement data is obtained, and the viewing angle change analysis of the constructed 3D space geometric model is performed based on these data. By quantifying the light change illusion caused by the movement of the viewing angle, the system evaluates the gradient of parallax and focus changes under different viewing angles. These data help the system understand the impact of different viewing angle switching on the user's visual perception, especially the changes in parallax and focus, and then evaluate the discomfort and visual fatigue caused. Based on the parallax focus gradient loss data, the system further evaluates the visual perception fatigue. This process analyzes the user's visual feedback at different viewing angles to detect the discomfort symptoms caused, such as dizziness or eye fatigue. According to the evaluation results, the system designed a light-variable control mechanism to dynamically adjust the changes in light to reduce the visual discomfort caused by parallax, optimize the user's visual experience, and improve the comfort of watching 3D images. The strategy gradient algorithm is used to perform logical memory learning on the light-variable control mechanism. By continuously optimizing the control strategy, the system can intelligently adjust the image display parameters under different user perspective changes to ensure the best visual effect. At the same time, this control logic is converted into automatic control firmware and embedded in the image control processor of the naked-eye 3D screen. It can automatically adjust the 3D image display parameters in real time without user intervention, thereby improving the stability of the visual effect and the comfort of the user experience. Therefore, the present invention is an optimization process made to a traditional 3D image display method, which solves the problems of low accuracy in visual fatigue analysis and large control accuracy error in a traditional 3D image display method, improves the accuracy of visual fatigue analysis, and reduces the control accuracy error.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: Acquire multimedia video playback image;

[0012] Step S12: performing image edge optimization processing on the multimedia video playback image to obtain video playback image edge optimization data;

[0013] Step S13: marking the background reference object-foreground subject plane coordinates of the video playback image edge optimization data to obtain the background reference object-foreground subject plane coordinates;

[0014] Step S14: constructing a background-foreground subject 3D space geometry model for the edge optimization data of the video playback image according to the background reference object-foreground subject plane coordinates to obtain a background-foreground subject 3D space geometry model.

[0015] The present invention first extracts image data from the video playback source. The core of this step is to capture high-quality image frames and provide original visual information for subsequent processing. By ensuring the clarity and accuracy of the image, the system can provide accurate input data for 3D space modeling. This step lays the foundation for subsequent image optimization, space modeling and visual effect improvement, ensuring that the system can obtain accurate scene information in order to handle complex visual tasks. The edge optimization processing of the video playback image is performed. This operation enhances the edge details of the image, improves the resolution and clarity of the image, and especially strengthens the foreground and background in complex scenes. The optimization of the image edge helps to improve the overall expression and depth of the image, making the image more accurate in subsequent space modeling, and providing higher quality data support for building a clear 3D geometric model. The optimized image is helpful for subsequent visual effect analysis and adjustment. By analyzing the optimized image, the plane coordinates between the background reference and the foreground subject (i.e., the theme element in the image, the character protagonist, etc.) are marked. This operation can accurately extract each element in the scene, especially the key boundaries and references separating the background and the foreground. By marking the plane coordinates, the system can clarify the spatial layout in the image, determine the relative positions of various objects, and provide an accurate reference frame for 3D space modeling. This process improves the accuracy of subsequent geometric model construction. Based on the plane coordinates of the marked background reference objects and foreground subjects, a 3D spatial geometric model of the background and foreground subjects of the video playback image is constructed. By combining the depth information and coordinate markings of the image, the system can accurately restore the three-dimensional structure of the scene. This process makes the image's expressiveness not limited to two-dimensional perspective, but extends to three-dimensional space, thereby providing sufficient spatial data support for subsequent perspective adjustment, image enhancement, and visual effect optimization.

[0016] Preferably, step S2 comprises the following steps:

[0017] Step S21: Acquire user perspective movement data;

[0018] Step S22: Based on the user perspective movement data, the focus area of ​​the background-foreground subject 3D space geometric model between different perspectives is identified to obtain the background-foreground subject perspective focus area data;

[0019] Step S23: quantifying the perspective movement light change illusion of the background-foreground subject perspective focus area data according to the user perspective movement data to obtain perspective movement light change illusion quantification data;

[0020] Step S24: performing parallax focus gradient loss evaluation on the background-foreground subject perspective focus area data according to the perspective movement light change illusion quantification data to obtain parallax focus gradient loss data.

[0021] The present invention obtains the user's perspective movement data in real time through sensors, cameras or user input devices. These data reflect the movement trajectory and direction changes of the user's eyes or head when watching videos or using 3D systems. Acquiring perspective data is crucial for subsequent steps because it provides dynamic information about the user's interaction with the image, which can help the system adapt to the user's viewing habits, thereby accurately adjusting the display effect of the image and providing a personalized visual experience. Based on the acquired user perspective movement data, the focus areas in the background and foreground subjects are analyzed and identified. These areas are the parts where the user's visual focus is located, and usually include detailed or eye-catching elements. Through this identification, the system can optimize the focus areas under different perspectives to ensure that these areas remain clear when the user's perspective moves and present a suitable sense of depth in the 3D space. This process helps to improve the accuracy of the visual focus, thereby enhancing the user's viewing experience. Based on the user's perspective movement data and the perspective focus area data, the illusion caused by the change of light in the image is quantified. As the user's perspective changes, the projection path of the light and the light and shadow effects under the perspective will also change, thereby affecting visual perception. By quantifying this illusion of light change, the system can evaluate and identify visual discomfort under different viewing angles, and provide a quantitative basis for subsequent optimization and adjustment. This step ensures that the system can accurately grasp the user's visual feedback and make corresponding adjustments. Based on the quantitative data of the illusion of light change caused by the movement of the perspective, the parallax focus gradient loss of the perspective focus area is evaluated. The parallax focus gradient loss is a key parameter to measure the depth difference and focus effect change between the background and foreground subjects during the movement of the user's perspective. Through this evaluation, the system can detect visual discomfort or focusing problems caused by changes in perspective, and provide a scientific basis for adjusting the image display. The evaluation results help the system optimize the sense of depth in the focus area, making the user's visual experience more natural and comfortable at different perspectives.

[0022] Preferably, step S23 includes the following steps:

[0023] Step S231: simulating and calculating the focus area perspective movement offset of the background-foreground subject perspective focus area data according to the user perspective movement data to obtain the focus area perspective movement offset;

[0024] Step S232: calculating the horizontal section offset angle of the focus area viewing angle movement offset to obtain horizontal section offset angle data;

[0025] Step S233: performing a light scattering geometric color wheel illusion evaluation on the background-foreground subject perspective focus area data according to the focus area perspective movement offset and the horizontal section offset angle data to obtain geometric color wheel illusion data;

[0026] Step S234: quantifying the perspective movement and light change illusion of the background-foreground subject perspective focus area data according to the geometric color wheel illusion data to obtain perspective movement and light change illusion quantification data.

[0027] The present invention simulates and calculates the offset of the perspective focus area in the background-foreground subject according to the movement data of the user's perspective. This calculation determines the position offset of the focus area relative to the initial perspective by accurately analyzing the dynamic changes of the user's perspective, ensuring that the system can track the user's perspective changes in real time. This process is crucial for subsequent light changes and visual effect optimization because it provides basic data for evaluating and correcting image performance under different perspectives. The calculated focus area perspective movement offset is further analyzed to calculate the horizontal section offset angle, which describes the displacement change of the focus area along the horizontal plane during the user's perspective movement. Through this calculation, the system can more accurately understand the spatial change of the focus area, thereby evaluating the image presentation effect under different perspectives. This step ensures the spatial sense and realism of the image, and can smoothly transition when the user's perspective changes, reducing visual illusions. Based on the focus area perspective movement offset and the horizontal section offset angle data, the light scattering effect in the image is evaluated for geometric color wheel illusion. The movement of the perspective not only changes the focus of the image, but also affects the propagation path and color performance of the light, resulting in a visual illusion that affects the user's perceived depth and details. By evaluating the geometric color wheel illusion, the system can analyze the impact of this illusion on visual effects and optimize it in a targeted manner to ensure that the color and brightness of the image remain consistent at different viewing angles, thereby improving the naturalness of the visual effect. Based on the geometric color wheel illusion data, the illusion of perspective movement and light change in the background-foreground subject perspective focus area is further quantified. This quantification process measures the degree of illusion of light, color, and depth during the perspective change in a numerical way. The system can fine-tune the image based on this quantitative data, optimize the user's visual experience, ensure smooth image transitions at different viewing angles, and reduce visual fatigue and illusions. Ultimately, this step can effectively enhance the immersion and authenticity of the image, allowing users to experience a higher quality visual experience during the viewing process.

[0028] Preferably, step S233 includes the following steps:

[0029] Performing background-foreground subject geometric morphology analysis on the background-foreground subject perspective focus area data to obtain background geometric morphology data and foreground subject geometric morphology data respectively;

[0030] Perform geometric surface normal vector interleaving calculation on the background geometric morphology data and the foreground subject geometric morphology data to obtain background geometric morphology normal vector interleaving data and foreground subject geometric morphology normal vector interleaving data respectively;

[0031] Perform background-foreground subject light scattering spectral interference calculation on the background geometric morphology normal vector interlaced data and the foreground subject geometric morphology normal vector interlaced data to obtain global light scattering spectral interference data;

[0032] Perform spatial dispersion distribution analysis based on global light scattering spectral interference data to obtain spatial dispersion distribution data;

[0033] Performing a perspective shift dispersion anisotropy analysis on the spatial dispersion distribution data according to the perspective shift offset of the focus area and the horizontal cross-section offset angle data to obtain perspective shift dispersion anisotropy data;

[0034] Performing a perspective shift dispersion overlap analysis on the perspective shift dispersion anisotropy data to obtain perspective shift dispersion overlap data;

[0035] The light scattering geometric color wheel illusion is evaluated based on the viewing angle shift dispersion anisotropy data and the viewing angle shift dispersion overlap data to obtain the geometric color wheel illusion data.

[0036] The present invention first performs geometric analysis on the perspective focus area data of the background and foreground subjects, and extracts the background geometric data and the foreground subject geometric data respectively. This analysis provides a basis for subsequent light scattering and illusion calculations, because different geometric forms have different effects on the reflection, refraction and scattering of light. By processing the geometric forms of the background and foreground separately, the system can finely capture their respective structural features and provide more accurate geometric basic data for the calculation of light interaction. This step ensures the accurate performance of the image at different viewing angles and lays a solid foundation for the evaluation of visual illusions. The geometric surface normal vectors are interlaced for the background geometric data and the foreground subject geometric data to obtain the background geometric normal vector interlaced data and the foreground subject geometric normal vector interlaced data respectively. The geometric surface normal vector is a key data used to describe the direction and normal direction of each point on the three-dimensional surface, which can affect the reflection and refraction behavior of light. The interlaced calculation simulates the light interaction between the background and foreground at different viewing angles, reflecting the impact of viewing angle changes on visual effects. This step provides accurate geometric information for subsequent light scattering and geometric illusion evaluation by accurately calculating the normal vector interlacing of different object surfaces, ensuring high-quality presentation of visual effects. The background geometric morphology normal vector interlacing data and the foreground subject geometric morphology normal vector interlacing data are used for light scattering spectral interference calculation to obtain global light scattering spectral interference data. Light will scatter and interfere when passing through different surfaces, angles and media, which will affect the visual color distribution, brightness change and depth perception. The global light scattering spectral interference calculation not only considers the light changes of a single object, but also simulates the complex light interaction process between the background and the foreground, which can more realistically reproduce the visual perception from the user's perspective, reduce illusions and visual discomfort. This step provides data support for the optimization of the light scattering effect, making the image color and brightness transition more natural and smooth under dynamic perspective. Based on the global light scattering spectral interference data, spatial dispersion distribution analysis is performed to obtain spatial dispersion distribution data. Dispersion is the color separation phenomenon caused by the wavelength and frequency of light when light passes through different media. In 3D images and virtual reality, dispersion affects the clarity, sharpness and depth perception of the image. By analyzing the spatial distribution of global light scattering and interference effects, the system can identify the dispersion characteristics under different viewing angles and light change conditions, and then adjust the color distribution in the image to make the colors more balanced and natural. This step effectively improves the accuracy and immersion of the visual effect and reduces visual fatigue. According to the focus area perspective movement offset and horizontal section offset angle data, the spatial dispersion distribution data is analyzed for perspective offset dispersion anisotropy to obtain perspective offset dispersion anisotropy data. As the viewing angle changes, the dispersion phenomenon in the image will change. Different angles lead to problems such as color offset and brightness change, affecting the natural sense of vision.By analyzing the dispersion anisotropy, the system can identify the different effects of dispersion on visual effects at different viewing angles, and then take corresponding adjustment measures to ensure that the color and brightness of the image are always consistent when the viewing angle changes. This step provides a fine basis for viewing angle adjustment for subsequent image optimization. The viewing angle offset dispersion anisotropy data is subjected to viewing angle offset dispersion overlap analysis to obtain viewing angle offset dispersion overlap data. This analysis further optimizes the visual effect of the image by evaluating the overlap of different color and brightness intervals when the viewing angle changes. Through this analysis, the system can identify the visual errors caused by dispersion overlap and take measures to adjust the transition of color and brightness so that when the user's viewing angle changes, the dispersion effect will not produce abrupt visual effects, ensuring that the user can smoothly and naturally transition to a new viewing angle. This step helps to reduce light distortion or color deviation while improving visual quality. Combined with the viewing angle offset dispersion anisotropy data and the viewing angle offset dispersion overlap data, the light scattering geometric color wheel illusion is evaluated to obtain the geometric color wheel illusion data. The geometric color wheel illusion refers to the visual illusion of color and geometric structure caused by light scattering, dispersion and perspective changes. This assessment helps the system identify which areas produce unrealistic visual effects and correct them by comprehensively analyzing the impact of perspective changes on color, light and depth. Through this assessment, the system can provide more consistent and realistic visual effects at different perspectives, avoid visual distortion or discomfort, and ultimately provide users with a more immersive and natural visual experience.

[0037] Preferably, step S24 comprises the following steps:

[0038] Step S241: performing parallax depth level analysis on the background-foreground subject perspective focus area data according to the perspective movement light change illusion quantification data to obtain the focus area parallax depth level data;

[0039] Step S242: calculating the average depth difference between the layers of the focus area parallax depth level data to obtain the average parallax depth difference of the regional level;

[0040] Step S243: evaluating the focus point imaging clarity loss on the perspective movement light change illusion quantification data according to the regional level parallax depth average difference to obtain the focus point imaging clarity loss data;

[0041] Step S244: performing parallax focus gradient loss evaluation based on the focus point imaging clarity loss data and the regional level parallax depth average difference to obtain parallax focus gradient loss data.

[0042] The present invention can accurately quantify the illusion effect of perspective movement on light changes by analyzing the parallax depth level of the background and foreground subject perspective focus areas, thereby providing more accurate depth level data for subsequent imaging quality analysis. This analysis process can effectively reveal the parallax difference between the background and foreground under different perspectives, thereby providing a scientific basis for the adjustment of the focal point and the optimization of image clarity. By calculating the depth mean difference of the parallax depth level data of the focus area, the difference between different depth levels in the area can be quantified, thereby evaluating the uniformity of the overall parallax. This calculation can help identify the unevenness of the depth level changes during the imaging process, thereby providing support for optimizing the focal length and visual effects of the image, and reducing visual discomfort caused by uneven parallax. After obtaining the regional level parallax depth mean difference, the focal point imaging clarity loss is further evaluated for the illusion of perspective movement and light changes, and the imaging clarity loss of the focal point can be quantified. This process enables the technology to accurately evaluate the degree of loss of imaging clarity under different focal length conditions, thereby providing a quantitative basis for improving the clarity and visual effects of the focal point, thereby improving image quality. Based on the imaging clarity loss data of the focal point and the average parallax depth difference of the regional level, the parallax focus gradient loss evaluation is performed, which can more comprehensively evaluate the imaging quality loss under complex viewing angles. This evaluation helps to further optimize the focal length focus and depth adjustment of the image and improve the accuracy of the imaging system. Especially when it is necessary to deal with complex background and foreground layers, it can effectively reduce the visual distortion and unclearness in the imaging, thereby improving the overall visual experience.

[0043] Preferably, step S3 comprises the following steps:

[0044] Step S31: performing a focus area spatial deformation analysis on the background-foreground subject 3D space geometric model based on the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain the focus area spatial deformation data;

[0045] Step S32: evaluating the visual perception fatigue of the focus area spatial deformation data according to the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain visual perception fatigue data;

[0046] Step S33: Designing a light variable control mechanism based on the visual perception fatigue data and the visual angle movement light change illusion quantification data and the parallax focus gradient loss data to obtain a light variable control mechanism.

[0047] The present invention first analyzes the 3D spatial geometric model of the background and foreground based on the quantified data of the illusion of light changes caused by the movement of the viewing angle and the parallax focus gradient loss data to identify the spatial deformation of the focus area. This analysis helps to accurately depict the spatial changes of objects and scenes in the focus area, thereby providing data support for the subsequent optimization of visual effects. This step can effectively evaluate the degree of distortion and dislocation of spatial geometry at different viewing angles, and provide an important basis for optimizing display effects and visual perception. Combined with the quantified data of the illusion of light changes caused by the movement of the viewing angle and the parallax focus gradient loss data, the spatial deformation data of the focus area is further analyzed to evaluate the visual perception fatigue caused. By quantifying the degree of visual fatigue, it is possible to identify the fatigue problems that occur in users under specific viewing angles or display settings. This evaluation provides important feedback for the subsequent control mechanism, ensuring that the technical solution can be actively adjusted to reduce visual fatigue, thereby improving visual comfort and user experience. Based on the visual perception fatigue data obtained in the first two steps, a light variable control mechanism is designed in this step to adjust light changes and viewing angle settings. This control mechanism can automatically adjust the changes in the display light source during the visual experience, optimize visual perception, reduce visual fatigue, and achieve a more comfortable visual effect by adjusting the viewing angle and light changes. This design can automatically adjust parameters according to different fatigue feedback, thereby providing optimal visual comfort in different usage scenarios and ensuring that users do not feel uncomfortable during long-term use.

[0048] Preferably, step S33 includes the following steps:

[0049] Step S331: performing focus drift effect analysis on the perspective movement light change illusion quantification data and the parallax focus gradient loss data based on the visual perception fatigue data to obtain focus drift effect data;

[0050] Step S332: performing color temperature-brightness coupling adjustment analysis on the focus drift effect data to obtain color temperature-brightness coupling adjustment data;

[0051] Step S333: performing local light field rendering deviation correction on the visual angle movement light change illusion quantization data according to the color temperature-brightness coupling adjustment data to obtain local light field rendering deviation correction data;

[0052] Step S334: performing focus space light correction on the focus drift effect data based on the local light field rendering deviation correction data to obtain focus space light correction data;

[0053] Step S335: Designing a light variable control mechanism according to the local light field rendering deviation correction data and the focus space light correction data to obtain a light variable control mechanism.

[0054] The present invention is based on visual perception fatigue data, and quantifies the focus drift effect by analyzing the perspective movement light change illusion quantification data and parallax focus gradient loss data. The focus drift effect refers to the phenomenon of the focus position shifting under different perspectives and light conditions, resulting in image blur or distortion. By accurately analyzing the focus drift effect, it is possible to identify under what circumstances the focus drifts, thereby providing data support for subsequent regulation and optimization, which is helpful to reduce visual fatigue and improve image clarity and stability. By analyzing the focus drift effect data, the coupling adjustment link between color temperature and brightness is entered. In this step, the color temperature and brightness are optimized and adjusted in combination with the characteristics of the focus drift effect, aiming to reduce visual discomfort and fatigue by adjusting the color temperature and brightness relationship of the light source. Through the reasonable coupling of color temperature and brightness, the overall comfort of the image can be improved, and the balance of visual effects under different display conditions can be ensured, so that users can maintain a high visual comfort when using the display device for a long time, and reduce fatigue caused by light changes. According to the color temperature-brightness coupling adjustment data, the local light field rendering deviation is corrected for the perspective movement light change illusion quantification data. Local light field rendering deviation refers to the phenomenon of inaccurate or inconsistent light field rendering in local areas under different light conditions and viewing angles. Through precise correction, problems such as image blur and color inconsistency caused by light field deviation can be effectively avoided. This step optimizes light field rendering, enhances the visual consistency and clarity of the image, and improves the user's visual experience. By using the local light field rendering deviation correction data, the focus drift effect data is further corrected for spatial light. Focus space light correction mainly adjusts the direction and focus point of the light to ensure that the focus position in space is accurate and clear. This correction step aims to eliminate the visual discomfort and image distortion caused by focus drift or light field deviation, so that the display device can maintain clear and stable imaging effects under different viewing angles and environments, further improving visual comfort and image quality. Finally, combined with the local light field rendering deviation correction data and the focus space light correction data, this step designs a light variable control mechanism. This mechanism can adjust light and display parameters in real time in a dynamic environment to cope with problems such as different viewing angles, light changes and focus drift. By intelligently controlling light and focus, it is ensured that the display device always maintains the best visual effect during long-term use, thereby reducing visual fatigue and improving the overall visual experience. This mechanism provides users with a more comfortable and personalized audio-visual experience while optimizing the display effect of the device.

[0055] Preferably, step S334 includes the following steps:

[0056] Based on the local light field rendering deviation correction data, the focus drift effect data is evaluated for the light field intensity distribution difference between different focus drift directions to obtain the light field intensity distribution difference data;

[0057] Decomposing the light field offset vectors between different focus drift directions on the light field intensity distribution difference data to obtain the light field offset vectors;

[0058] Reconstructing the focus error vector of the focus drift effect data according to the light field offset vector to obtain focus error vector reconstruction data;

[0059] Based on the focus error vector reconstruction data and the local light field rendering deviation correction data, spatial light scattering interference analysis is performed to obtain focus light scattering interference data;

[0060] Focus space light correction is performed on the focus drift effect data according to the focus light scattering interference data and the focus error vector reconstruction data to obtain focus space light correction data.

[0061] The present invention evaluates the difference in light field intensity distribution between different focus drift directions based on local light field rendering deviation correction data. This evaluation aims to quantify the change in light field intensity caused by focus drift in different directions, and then reveal the unevenness of light field distribution caused by viewing angle change or focus drift. By evaluating the difference in light field intensity, key information can be provided for subsequent optical correction, ensuring the stability of the focus position and the consistency of image quality, and reducing visual discomfort and blur effects. The light field intensity distribution difference data is decomposed to obtain the light field offset vector between different focus drift directions. The light field offset vector represents the direction and amplitude of the light offset in the light field due to focus drift or viewing angle change. Through the precise decomposition of the offset vector, the specific situation of the light field offset in different directions can be fully understood, and the necessary numerical basis can be provided for the next step of error vector reconstruction and spatial light correction. This analysis can help optimize the optical performance of the display device in various environments and improve the stability of the visual effect. Based on the light field offset vector, the focus error vector is reconstructed to obtain the focus error vector reconstruction data. The purpose of focus error vector reconstruction is to more accurately restore the error source in the focus drift effect. After reconstruction, more precise focus position data can be obtained. This data reveals the optical error caused by focus drift or light field offset and provides a basis for subsequent fine control. Through this process, the focus position can be adjusted in a targeted manner to ensure the accuracy and stability of the display effect. Based on the focus error vector reconstruction data and the local light field rendering deviation correction data, spatial light scattering interference analysis is performed. This analysis simulates and evaluates the interference effect in the light field by considering the influence of focus error on light scattering. Interference phenomena can cause optical distortion in the image and affect the visual quality. Through spatial light scattering interference analysis, the control of the light field can be further refined, and image problems caused by light scattering or interference can be accurately identified and corrected, laying the foundation for focus space light correction and reducing visual interference. Finally, based on the focus light scattering interference data and the focus error vector reconstruction data, focus space light correction is performed to obtain focus space light correction data. Through this correction process, the focus position and the distribution of the light field are accurately adjusted to ensure that the focus is always in the best position under different viewing angles and display conditions to avoid blur or visual fatigue. Focus spatial light correction can effectively eliminate inconsistencies and errors in the light field, providing clear and stable image display effects. Ultimately, this step optimizes the entire display process and improves the user's visual comfort and viewing experience.

[0062] Preferably, a naked-eye 3D screen is also provided, comprising a naked-eye 3D screen image control processor, which is used to execute the 3D image display method as described above, and the naked-eye 3D screen comprises:

[0063] The 3D space geometry model building module is used to obtain the multimedia video playback image; construct the background-foreground subject 3D space geometry model of the multimedia video playback image to obtain the background-foreground subject 3D space geometry model;

[0064] The light change illusion analysis module is used to obtain the user's perspective movement data; according to the user's perspective movement data, the perspective movement light change illusion is quantified for the background-foreground subject 3D space geometric model to obtain the perspective movement light change illusion quantification data; according to the perspective movement light change illusion quantification data, the parallax focus gradient loss is evaluated to obtain the parallax focus gradient loss data;

[0065] A light variable control mechanism design module is used to evaluate visual perception fatigue according to parallax focus gradient loss data to obtain visual perception fatigue data; based on the visual perception fatigue data, the light variable control mechanism is designed for the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain the light variable control mechanism;

[0066] The firmware design module is used to perform logic memory learning on the optically variable control mechanism based on the policy gradient algorithm to obtain the optically variable control logic memory data; perform automated control firmware design based on the optically variable control logic memory data to obtain the optically variable control logic firmware, and embed the optically variable control logic firmware into the naked-eye 3D screen image control processor to execute the 3D image display method.

[0067] Beneficial effects of the present invention:

[0068] By acquiring images played by multimedia videos, computer vision and deep learning algorithms are used to extract the spatial information of the background and foreground subjects (i.e., the theme elements in the image, the main characters, etc.), and perform three-dimensional geometric modeling. This process can convert the image content in the video into a 3D data structure that can be deeply analyzed and processed, thereby providing a basic geometric framework for subsequent operations such as perspective adjustment and light changes, making the subsequent visual effects more accurate and natural. The user's perspective movement data is obtained, and the perspective change analysis of the constructed 3D space geometric model is performed based on this data. By quantifying the light change illusion caused by the movement of the perspective, the system evaluates the gradient of parallax and focus changes under different perspectives. These data help the system understand the impact of different perspective switching on the user's visual perception, especially the changes in parallax and focus, and then evaluate the discomfort and visual fatigue caused. Based on the parallax focus gradient loss data, the system further evaluates the visual perception fatigue. This process analyzes the user's visual feedback under different perspectives to detect the discomfort symptoms caused, such as dizziness or eye fatigue. According to the evaluation results, the system designed a light-variable control mechanism to dynamically adjust the changes in light to reduce the visual discomfort caused by parallax, optimize the user's visual experience, and improve the comfort of watching 3D images. The strategy gradient algorithm is used to perform logical memory learning on the light-variable control mechanism. By continuously optimizing the control strategy, the system can intelligently adjust the image display parameters under different user perspective changes to ensure the best visual effect. At the same time, this control logic is converted into automatic control firmware and embedded in the image control processor of the naked-eye 3D screen. It can automatically adjust the 3D image display parameters in real time without user intervention, thereby improving the stability of the visual effect and the comfort of the user experience. Therefore, the present invention is an optimization process made to a traditional 3D image display method, which solves the problems of low accuracy in visual fatigue analysis and large control accuracy error in a traditional 3D image display method, improves the accuracy of visual fatigue analysis, and reduces the control accuracy error. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 A schematic diagram of a process flow of a 3D image display method;

[0070] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0071] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0072] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0073] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0074] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0075] It should be understood that although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit, and the term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0076] To achieve this, please refer to Figures 1 to 3 , a 3D image display method, the method comprising the following steps:

[0077] Step S1: Acquire a multimedia video playback image; construct a background-foreground subject 3D space geometry model for the multimedia video playback image to obtain a background-foreground subject 3D space geometry model;

[0078] Step S2: obtaining user perspective movement data; performing perspective movement light change illusion quantification on the background-foreground subject 3D space geometric model according to the user perspective movement data to obtain perspective movement light change illusion quantification data; performing parallax focus gradient loss evaluation according to the perspective movement light change illusion quantification data to obtain parallax focus gradient loss data;

[0079] Step S3: evaluating visual perception fatigue according to the parallax focus gradient loss data to obtain visual perception fatigue data; designing a light variable control mechanism for the perspective movement light change illusion quantification data and the parallax focus gradient loss data based on the visual perception fatigue data to obtain a light variable control mechanism;

[0080] Step S4: Based on the policy gradient algorithm, logic memory learning is performed on the optically variable control mechanism to obtain optically variable control logic memory data; according to the optically variable control logic memory data, automated control firmware design is performed to obtain optically variable control logic firmware, and the optically variable control logic firmware is embedded into the naked-eye 3D screen image control processor to execute the 3D image display method.

[0081] In the embodiment of the present invention, reference Figure 1 FIG. 1 is a schematic diagram of a step flow of a 3D image display method of the present invention. In this example, the 3D image display method includes the following steps:

[0082] Step S1: Acquire a multimedia video playback image; construct a background-foreground subject 3D space geometry model for the multimedia video playback image to obtain a background-foreground subject 3D space geometry model;

[0083] In an embodiment of the present invention, when extracting frame data from a video stream, the video signal processing module is first connected through a physical interface to obtain high frame rate frame-by-frame image information from the video source to ensure the temporal integrity and resolution accuracy of the extracted image. An image segmentation algorithm, such as a segmentation method based on active contours, is used to perform pixel-level analysis on the video frame to separate dynamic objects from static background areas. The segmentation result corrects the discontinuous boundaries through morphological operations to ensure the segmentation quality. Through point cloud reconstruction technology and multi-view projection method, multiple frame images are combined with the camera model after internal reference correction, and a triangulation algorithm is used to generate a spatial depth map. Using depth information, the segmented foreground subject and background area are respectively constructed into 3D geometric models through the voxel modeling method to refine the spatial shape. During the model optimization process, the Laplacian smoothing algorithm is used to eliminate surface noise in the reconstruction process, and the surface smoothness and authenticity are enhanced through normal vector resampling to generate a background-foreground subject 3D spatial geometric model.

[0084] Step S2: obtaining user perspective movement data; performing perspective movement light change illusion quantification on the background-foreground subject 3D space geometric model according to the user perspective movement data to obtain perspective movement light change illusion quantification data; performing parallax focus gradient loss evaluation according to the perspective movement light change illusion quantification data to obtain parallax focus gradient loss data;

[0085] In an embodiment of the present invention, the motion data of the user's head or eyes is captured in real time by an embedded sensor (such as an IMU unit or an eye tracker). The inertial parameters recorded by the sensor include angular velocity, linear acceleration, etc., and the perspective movement trajectory is dynamically reconstructed in combination with an attitude solution algorithm (such as quaternion attitude solution) to generate three-dimensional perspective movement data. Based on the user's perspective data, the light change is modeled as the dynamic input of the ray tracing algorithm (Ray Tracing). The algorithm simulates the refraction, reflection and shadow distribution of light based on the light source and the surface normal vector of the object, and calculates the change of light distribution in real time. Further, the spatial illusion caused by the light change is mathematically modeled using an illusion evaluation quantification model, and the illusion parameters include variables such as the light source offset angle, the light intensity gradient, and the line of sight focus error. The error propagation method is used to comprehensively evaluate the impact of light changes on the perspective perception, and the perspective movement light change illusion quantification data is obtained. Subsequently, based on the above quantitative data, the degree of depth misalignment of the user's line of sight focus is analyzed using the parallax evaluation function, and the focus error accumulation at different parallax positions is calculated using the gradient descent method to generate parallax focus gradient loss data.

[0086] Step S3: evaluating visual perception fatigue according to the parallax focus gradient loss data to obtain visual perception fatigue data; designing a light variable control mechanism for the perspective movement light change illusion quantification data and the parallax focus gradient loss data based on the visual perception fatigue data to obtain a light variable control mechanism;

[0087] In an embodiment of the present invention, based on the parallax focus gradient loss data, a visual fatigue assessment framework is adopted to evaluate the change in the user's visual load by calculating the ratio of the cumulative parallax error and the number of frequent focus adjustments. The spectral distribution of the parallax fluctuation is analyzed using Fourier transform, the parallax components of high-frequency changes are extracted, and the cumulative fatigue index is calculated. In combination with the visual fatigue assessment results, a light-variable control mechanism is designed through a parameter selection method based on an optimization algorithm (such as multi-objective particle swarm optimization). The mechanism consists of two parts: one is to adjust the brightness gradient of the display light source, and the other is to adjust the directional distribution of light reflection. The numerical integration technology is used to simulate the light source intensity distribution and generate a dynamic light source adjustment function; the direction field of the reflected light is calculated through the principle of geometric optics, and the light distribution control function is generated in combination with the gradient change strategy to complete the design of the light-variable control mechanism.

[0088] Step S4: Based on the policy gradient algorithm, logic memory learning is performed on the optically variable control mechanism to obtain optically variable control logic memory data; according to the optically variable control logic memory data, automated control firmware design is performed to obtain optically variable control logic firmware, and the optically variable control logic firmware is embedded into the naked-eye 3D screen image control processor to execute the 3D image display method.

[0089] In an embodiment of the present invention, the parameter optimization of the light-variable control mechanism is trained using a policy gradient algorithm. The algorithm takes light source control parameters and user feedback data (such as visual fatigue change indicators) as input, and optimizes the performance of the control strategy through policy iteration and gradient ascent methods. In the specific implementation, the light source brightness and reflection direction are used as the main optimization variables, and the policy parameters are iteratively updated until they converge to the optimal state, generating light-variable control logic memory data. The logic memory data is integrated into the hardware description language (HDL) for the development of embedded control firmware. The firmware development process includes: generating dynamic control logic according to the control strategy; using FPGA programming technology to implement the logic circuit in the naked-eye 3D screen image control processor; embedding the firmware into the controller through the physical interface, and performing circuit timing simulation tests to verify the firmware's response speed and stability. Finally, the design and embedding of the light-variable control logic firmware is completed to achieve automatic control of 3D image display.

[0090] Preferably, step S1 comprises the following steps:

[0091] Step S11: Acquire multimedia video playback image;

[0092] Step S12: performing image edge optimization processing on the multimedia video playback image to obtain video playback image edge optimization data;

[0093] Step S13: marking the background reference object-foreground subject plane coordinates of the video playback image edge optimization data to obtain the background reference object-foreground subject plane coordinates;

[0094] Step S14: constructing a background-foreground subject 3D space geometry model for the edge optimization data of the video playback image according to the background reference object-foreground subject plane coordinates to obtain a background-foreground subject 3D space geometry model.

[0095] In an embodiment of the present invention, the operation of extracting a video playback image from a multimedia data source requires first connecting a physical interface, such as an HDMI input signal channel or a high-bandwidth digital video interface. When acquiring a video stream, the input video signal is parsed using decoder hardware and converted into a frame sequence. The extraction of the frame sequence should maintain the order integrity on the time axis and verify the image resolution to ensure that it meets the requirements of subsequent processing. A video compression format based on DCT (discrete cosine transform) is used to restore high-fidelity video frame images through an inverse vector quantization method. After the extraction is completed, the frame sequence is stored in a high-performance memory for subsequent processing. For the extracted video playback image, edge optimization processing is applied to enhance the clarity of the image contour. First, the Canny edge detection algorithm is used to extract the preliminary edge information of the image, and the contour points are obtained by calculating the image pixel gradient. Then, the breakpoints of the edge lines are repaired using morphological dilation and erosion operations to make the edge information more continuous. To reduce noise interference, bilateral filtering is performed on the image to simultaneously retain edge information and remove noise in flat areas. In addition, in order to enhance the edge details, the Laplacian of Gaussian (LoG) operator can be used to perform secondary edge detection. The edge optimization data of the video playback image generated after processing is stored in the buffer as the input of the subsequent marking operation. The edge optimization data of the video playback image is used to achieve the preliminary separation of the background reference and the foreground subject through the image segmentation algorithm. The image segmentation adopts the region growing method, starting from the selected seed pixel, and recursively marks the region range based on the grayscale similarity of adjacent pixels. After the segmentation is completed, the centroid calculation algorithm is used to obtain the geometric center point coordinates of the foreground subject and the background reference. Then, for the background area, the point cloud registration algorithm (such as the ICP algorithm) is used to generate its two-dimensional plane coordinate description; for the foreground area, the pixel projection technology is used to determine its coordinate boundary on the video image plane. In order to ensure the accuracy of the marking result, the shape matching algorithm is introduced in the marking process, and the actual boundary contour is compared and calibrated with the standard contour template, and finally the background reference-foreground subject plane coordinates are generated, where the foreground subject refers to the dynamic character or the visual focus displayed by the image; according to the background reference-foreground subject plane coordinates, the plane coordinates are converted into three-dimensional coordinates through the triangulation algorithm. The specific operation is to perform stereo matching on the corresponding points of each pair of images, calculate the depth difference between the point pairs using the principle of multi-view geometry, and generate preliminary three-dimensional point cloud data. After generating the point cloud, the Delaunay triangulation algorithm is used to construct the 3D surface mesh structure. After the mesh is constructed, the voxel segmentation method is used to refine the model to eliminate abnormal points caused by uneven depth. Furthermore, through normal vector resampling and surface smoothing, the surface smoothness and model integrity are optimized, and finally the background-foreground subject 3D spatial geometric model is obtained and stored in the spatial geometry database for subsequent use.

[0096] Preferably, step S2 comprises the following steps:

[0097] Step S21: Acquire user perspective movement data;

[0098] Step S22: Based on the user perspective movement data, the focus area of ​​the background-foreground subject 3D space geometric model between different perspectives is identified to obtain the background-foreground subject perspective focus area data;

[0099] Step S23: quantifying the perspective movement light change illusion of the background-foreground subject perspective focus area data according to the user perspective movement data to obtain perspective movement light change illusion quantification data;

[0100] Step S24: performing parallax focus gradient loss evaluation on the background-foreground subject perspective focus area data according to the perspective movement light change illusion quantification data to obtain parallax focus gradient loss data.

[0101] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0102] Step S21: Acquire user perspective movement data;

[0103] In an embodiment of the present invention, the process of acquiring user perspective movement data needs to combine sensor hardware and data acquisition algorithms. First, an infrared eye tracking sensor and a high-precision accelerometer are embedded in the naked-eye 3D display device, and the position of the user's gaze point is captured in real time through eye tracking technology. The Kalman filter algorithm is used to dynamically predict and suppress noise on the captured data to ensure the accuracy of the line of sight movement trajectory. Secondly, the inclination angle and rotation direction of the user's head are obtained in combination with the accelerometer data, and these data are converted into a rotation matrix representation in three-dimensional space using the Euler angle formula. By combining the line of sight gaze point and the head rotation angle, the user's dynamic perspective matrix data is generated. In order to ensure the synchronization of data in time, a timestamp is introduced to uniformly mark all data, and the perspective movement data of the interval frames is compensated by a linear interpolation method, and finally the complete user perspective movement data is generated.

[0104] Step S22: Based on the user perspective movement data, the focus area of ​​the background-foreground subject 3D space geometric model between different perspectives is identified to obtain the background-foreground subject perspective focus area data;

[0105] In an embodiment of the present invention, the user's perspective movement data is used to dynamically analyze the background-foreground subject 3D space geometric model, and the multi-perspective ray projection method is used to identify the focus area under different perspectives. In the specific implementation process, the position of the viewpoint is first determined according to the user's perspective matrix, and the background-foreground subject 3D space geometric model is projected into the user's visual plane to generate a real-time two-dimensional projection image. Then, the image area on the visual plane is segmented by a focus detection algorithm. In this process, a focus calculation method based on Gaussian blur weights is used to identify the main focus area of ​​the user's line of sight. After the focus area is identified, the three-dimensional coordinate range of the focus area in the field of view is screened out by the Z-buffer depth test algorithm in combination with the depth data in the three-dimensional geometric model, and finally the background-foreground subject perspective focus area data is generated.

[0106] Step S23: quantifying the perspective movement light change illusion of the background-foreground subject perspective focus area data according to the user perspective movement data to obtain perspective movement light change illusion quantification data;

[0107] In an embodiment of the present invention, a ray tracing algorithm is used to simulate and analyze the light changes caused by the movement of the perspective according to the user's perspective movement data and the background-foreground subject perspective focus area data. First, a light refraction and reflection model based on the Fresnel equation is constructed to simulate the propagation path of light from the background-foreground subject surface to the user's viewpoint. During the simulation process, the interference effect of different material surfaces on the light propagation path is calculated, and the path deviation is quantified using the Monte Carlo random sampling technique. An illusion quantification index is introduced to calculate the illusion intensity by the ratio of the line of sight movement speed to the light path offset angle. Combined with the frequency analysis data of the user's perspective change, the discrete Fourier transform (DFT) is used to further decompose and reconstruct the frequency domain characteristics of the light illusion change, and finally generate the perspective movement light change illusion quantification data.

[0108] Step S24: performing parallax focus gradient loss evaluation on the background-foreground subject perspective focus area data according to the perspective movement light change illusion quantification data to obtain parallax focus gradient loss data.

[0109] In an embodiment of the present invention, the disparity gradient analysis method is used to evaluate the disparity focus gradient loss by using the quantified data of the illusion of light changes caused by the movement of the viewing angle and the background-foreground subject viewing angle focus area data. First, the disparity value change of each focus area during the viewing angle movement is obtained by the binocular disparity calculation formula, and the three-dimensional coordinate difference of the focus points of the two viewing angles is recorded. Then, the gradient direction and amplitude analysis are used to perform a gradient evaluation of the spatial distribution of the disparity value. The rate of change and fluctuation amplitude of the disparity between different focus areas are calculated by a gradient loss function (such as Laplacian gradient loss). In order to improve the stability of the evaluation, a Bayesian optimization algorithm is introduced to dynamically adjust the parameters in the evaluation process to eliminate the influence of outliers on the loss calculation. Finally, all disparity evaluation results are integrated to generate disparity focus gradient loss data to provide data support for the subsequent design of optical control mechanisms.

[0110] Preferably, step S23 includes the following steps:

[0111] Step S231: simulating and calculating the focus area perspective movement offset of the background-foreground subject perspective focus area data according to the user perspective movement data to obtain the focus area perspective movement offset;

[0112] Step S232: calculating the horizontal section offset angle of the focus area viewing angle movement offset to obtain horizontal section offset angle data;

[0113] Step S233: performing a light scattering geometric color wheel illusion evaluation on the background-foreground subject perspective focus area data according to the focus area perspective movement offset and the horizontal section offset angle data to obtain geometric color wheel illusion data;

[0114] Step S234: quantifying the perspective movement and light change illusion of the background-foreground subject perspective focus area data according to the geometric color wheel illusion data to obtain perspective movement and light change illusion quantification data.

[0115] In an embodiment of the present invention, the offset of the focus area caused by the perspective movement is calculated by combining the user perspective movement data and the background-foreground subject perspective focus area data. First, by extracting the time series information of the user perspective matrix, the dynamic change of the perspective is parsed into a combination of a translation vector and a rotation matrix. Subsequently, the background-foreground subject perspective focus area data is mapped to a new perspective coordinate system using a homogeneous coordinate transformation formula, thereby calculating the three-dimensional coordinate difference of each focus point before and after the perspective change. In the calculation process, an interpolation algorithm is introduced to smooth the perspective movement trajectory, eliminate the discrete jump phenomenon in the data, and ensure the continuity and accuracy of the calculation results. By performing time integration on the offset vector of each frame perspective change, the cumulative effect of the overall perspective offset is simulated to generate the perspective movement offset of the focus area. The perspective movement offset data of the focus area is used to calculate the offset angle on the horizontal section. In specific implementation, the perspective offset is first decomposed into a horizontal component and a vertical component, and all three-dimensional coordinate data are mapped to a horizontal plane by a projection method. The angle between the offset vector and the reference vector (such as the initial direction vector of the viewing angle) is calculated using the law of cosines. The time series changes of the angle are recorded by traversing the data at each time point. The angle calculation results are denoised and optimized by combining data smoothing algorithms such as the Savitzky-Golay filter, and the horizontal section offset angle data are finally generated. The light scattering geometric color wheel illusion of the background-foreground subject viewing angle focusing area is evaluated by combining the viewing angle movement offset of the focusing area and the horizontal section offset angle data. First, based on the theory of geometric optics, the changes in the light propagation path caused by the change in viewing angle are simulated, and the starting point, refraction point, reflection point and end point of each light are dynamically located by using the reverse ray tracing algorithm to obtain the scattering trajectory of each light. Subsequently, the light scattering results are mapped to the CIE color wheel model, and the mixing effect of light of different wavelengths is calculated according to the light path offset and angle change, and the law of color distribution change of visual perception is obtained. The periodic characteristics of color change are analyzed by discrete Fourier transform (DFT) to quantify the intensity of the geometric color wheel illusion. During the analysis process, a weight function is introduced to normalize the scattering path length and light intensity to ensure the uniformity and objectivity of the evaluation results. Finally, the geometric color wheel illusion data is output. The geometric color wheel illusion data and the background-foreground subject perspective focus area data are used to quantitatively analyze the illusion of light change due to perspective movement. First, by establishing a correlation matrix between light change and user perspective movement, the intensity of the geometric color wheel illusion is mapped to the perspective change frequency of different focus areas. The least squares fitting algorithm is used to fit the data to extract the dominant mode and trend of light change. Then, the light change is quantified into a specific illusion indicator through polarization angle change analysis and optical distortion coefficient calculation. Time domain signal analysis tools (such as wavelet transform) are introduced to perform time-frequency decomposition of the light change signal to capture the illusion peak area when the perspective moves rapidly.Combining these analysis results, quantitative data on the illusion of light changes caused by perspective movement are generated comprehensively, providing a quantitative basis for subsequent regulatory design.

[0116] Preferably, step S233 includes the following steps:

[0117] Performing background-foreground subject geometric morphology analysis on the background-foreground subject perspective focus area data to obtain background geometric morphology data and foreground subject geometric morphology data respectively;

[0118] Perform geometric surface normal vector interleaving calculation on the background geometric morphology data and the foreground subject geometric morphology data to obtain background geometric morphology normal vector interleaving data and foreground subject geometric morphology normal vector interleaving data respectively;

[0119] Perform background-foreground subject light scattering spectral interference calculation on the background geometric morphology normal vector interlaced data and the foreground subject geometric morphology normal vector interlaced data to obtain global light scattering spectral interference data;

[0120] Perform spatial dispersion distribution analysis based on global light scattering spectral interference data to obtain spatial dispersion distribution data;

[0121] Performing a perspective shift dispersion anisotropy analysis on the spatial dispersion distribution data according to the perspective shift offset of the focus area and the horizontal cross-section offset angle data to obtain perspective shift dispersion anisotropy data;

[0122] Performing a perspective shift dispersion overlap analysis on the perspective shift dispersion anisotropy data to obtain perspective shift dispersion overlap data;

[0123] The light scattering geometric color wheel illusion is evaluated based on the viewing angle shift dispersion anisotropy data and the viewing angle shift dispersion overlap data to obtain the geometric color wheel illusion data.

[0124] In the embodiment of the present invention, first, the three-dimensional geometric point cloud data in each focus area is extracted by combining the background-foreground subject perspective focus area data. The surface mesh model of the background and foreground subject is constructed by using the Delaunay triangulation algorithm by performing a neighborhood search on the point cloud data. The background geometric morphology analysis focuses on extracting geometric features on large static objects (such as the ground and buildings), while the foreground subject geometric morphology analysis focuses on the detailed features of dynamic or moving objects (such as people and vehicles). The noise in the point cloud is smoothed by the surface fitting method to eliminate the geometric distortion caused by the measurement error. At the same time, the Gaussian curvature calculation method is used to classify the morphological features of the geometric surfaces of the background and foreground subject. The area with positive Gaussian curvature is defined as a convex structure, the area with negative Gaussian curvature is defined as a concave structure, and the Gaussian curvature of the flat area is close to zero. Finally, the processed geometric morphology is divided into background geometric morphology data and foreground subject geometric morphology data, which are stored separately and used as input for subsequent calculations. For each triangular mesh face in the background geometric morphology data and the foreground subject geometric morphology data, its normal vector is calculated. The specific operation is to obtain the normal vector direction by solving the cross product between the coordinate points of the three vertices of each triangle. Subsequently, these normal vectors are used to construct a normal vector field, and the surface normal vectors of the entire geometric form are globally mapped. In the interlaced calculation process, the intersection area of ​​the background and foreground geometric form normal vector fields is selected as the key analysis object. The vector inner product method is used to calculate the angle between each pair of normal vectors, and the angle distribution in the interlaced area is further used to statistically summarize the degree of normal vector interlacing, and the background geometric form normal vector interlacing data and the foreground subject geometric form normal vector interlacing data are obtained respectively. Based on the background geometric form normal vector interlacing data and the foreground subject geometric form normal vector interlacing data, the scattering path of light and its interaction with the geometric surface are calculated. First, the propagation path of light is simulated using the theory of geometric optics, and the scattering behavior of each incident light is tracked with the background and foreground subject geometric surfaces as reflection boundaries. The calculation of the scattering path is based on the ray tracing algorithm, and the behavior of light in the interlaced area is simulated by the reflection law and the refraction law respectively. Then, the light scattering path is projected into the spectral space, and the intensity change of light in the spectroscopic effect is analyzed by Fourier transform. On this basis, combined with the theory of spectral interference, the interference intensity is accurately calculated using the multilayer film interference formula. By calculating the global interference intensity point by point and integrating the light scattering effect between the background and foreground subjects, the global light scattering spectral interference data is obtained as the final output.

[0125] Preferably, step S24 includes the following steps:

[0126] Step S241: performing parallax depth level analysis on the background-foreground subject perspective focus area data according to the perspective movement light change illusion quantification data to obtain the focus area parallax depth level data;

[0127] Step S242: calculating the average depth difference between the layers of the focus area parallax depth level data to obtain the average parallax depth difference of the regional level;

[0128] Step S243: evaluating the focus point imaging clarity loss on the perspective movement light change illusion quantification data according to the regional level parallax depth average difference to obtain the focus point imaging clarity loss data;

[0129] Step S244: performing parallax focus gradient loss evaluation based on the focus point imaging clarity loss data and the regional level parallax depth average difference to obtain parallax focus gradient loss data.

[0130] In the embodiment of the present invention, based on the global light scattering spectroscopic interference data, the angle offset and wavelength distribution of each light scattering path are first accurately analyzed. Using the spectral distribution characteristics, the light is segmented by wavelength and the trend of the change of its scattering intensity on the geometric surface is calculated. The degree of dispersion of light at different wavelengths is determined by the gradient distribution of the scattering intensity. In order to obtain the spatial distribution of dispersion, a three-dimensional interpolation algorithm is used to map the spectral data to a three-dimensional coordinate space to form a spatial dispersion map. The dispersion density of different regions is calculated by spatial frequency distribution analysis technology, and the periodic structure in the dispersion map is extracted by fast Fourier transform (FFT) in combination with the local change of the scattered light intensity to obtain the characteristic pattern of spatial dispersion distribution. Finally, it is sorted and summarized into spatial dispersion distribution data to provide input for the next step of analysis. Using the focus area perspective movement offset data, the movement trajectory of the perspective in space is simulated to extract the light scattering dispersion distribution changes at different positions during the movement of the perspective. Combined with the horizontal section offset angle data, the relationship between the perspective offset and the spatial dispersion distribution is determined. Through the vector rotation matrix, the deviation direction of the viewing angle movement is mathematically coupled with the local characteristics of the light dispersion, and the dispersion change trend at each viewing angle position is analyzed. The spatial anisotropy analysis method is used to quantify the directional change of the dispersion intensity. Through the main direction vector calculation method, the dispersion distribution in each direction is decomposed, and the degree of difference in different directions is calculated to form a complete viewing angle shift dispersion anisotropy characteristic matrix. The results are organized into viewing angle shift dispersion anisotropy data to evaluate the impact of viewing angle shift on the overall dispersion distribution. Based on the viewing angle shift dispersion anisotropy data, the dispersion characteristics of the overlapping area during the viewing angle movement are analyzed. By constructing a multi-view overlap matrix, the degree of dispersion overlap in different viewing angle areas during viewing angle movement is calculated. The convolution integral method is used to simulate multiple coverage of the same area by viewing angle movement, and the dispersion intensity superposition effect of the overlapping area is analyzed. Two-dimensional histogram statistics are performed on the dispersion overlap area, and the relationship data between the dispersion intensity and the number of overlaps are extracted respectively. Through the density estimation method, the dispersion degree change in the overlapping area is quantified to form a complete overlap analysis model. Finally, the viewing angle offset dispersion overlap data is calculated and used as the key input for the subsequent light scattering geometric color wheel illusion evaluation. Combining the viewing angle offset dispersion anisotropy data and the viewing angle offset dispersion overlap data, the global dispersion effect of different viewing angle offsets on light scattering is first analyzed. Using the principles of geometric optics and color theory, a geometric color wheel mapping of light scattering is constructed. Through multivariate nonlinear function fitting, the effects of dispersion anisotropy and dispersion overlap are combined into a comprehensive illusion index, and the calculated illusion index distribution is interpolated and fitted to generate a geometric color wheel illusion distribution map. Further, the color wheel's hue ring shift trend is analyzed to quantify the visual impact of the illusion, and finally the geometric color wheel illusion data is obtained as the key basis for subsequent 3D image display optimization.

[0131] Preferably, step S3 comprises the following steps:

[0132] Step S31: performing a focus area spatial deformation analysis on the background-foreground subject 3D space geometric model based on the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain the focus area spatial deformation data;

[0133] Step S32: evaluating the visual perception fatigue of the focus area spatial deformation data according to the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain visual perception fatigue data;

[0134] Step S33: Designing a light variable control mechanism based on the visual perception fatigue data and the visual angle movement light change illusion quantification data and the parallax focus gradient loss data to obtain a light variable control mechanism.

[0135] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0136] Step S31: performing a focus area spatial deformation analysis on the background-foreground subject 3D space geometric model based on the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain the focus area spatial deformation data;

[0137] In an embodiment of the present invention, the 3D spatial geometric model of the background-foreground subject is mapped and analyzed point by point by inputting the quantified data of the illusion of light changes caused by the movement of the viewing angle and the parallax focusing gradient loss data. First, the parallax-focus relationship mapping matrix of each pixel point in the three-dimensional space is established by using the quantified data of the illusion of light changes caused by the movement of the viewing angle. Combined with the parallax focusing gradient loss data, the curvature is extracted from the gradient change of the focus area in the mapping matrix to identify the potential deformation area. For the identified focus area, the three-dimensional Lagrange interpolation method is applied to refine the geometric morphology of the local area and calculate the change value of the surface tension during the deformation process. At the same time, the stepwise approximation algorithm is used to determine the spatial curvature change trend of the deformation area under the influence of parallax focusing. Finally, based on the deformation analysis results, the deformation data of the focus area is saved as the spatial deformation data of the focus area, providing accurate input for the subsequent visual perception fatigue assessment.

[0138] Step S32: evaluating the visual perception fatigue of the focus area spatial deformation data according to the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain visual perception fatigue data;

[0139] In an embodiment of the present invention, the spatial deformation data of the focus area is input into the visual perception fatigue assessment framework, and the correlation model between the three is established by combining the quantification data of the illusion of light change caused by the movement of the viewing angle and the parallax focus gradient loss data. The correlation between the visual fixation time and the deformation change frequency of the focus area is calculated through the theory of visual physiology. The local sensitivity analysis method is used to quantify the stimulation intensity of the deformation area to the visual system. In the specific implementation, the frequency domain analysis method based on Fourier transform is used to evaluate the visual perception fatigue, and the deformation data is converted into frequency domain characteristics to extract the influence of high-frequency components on visual fatigue. At the same time, the parallax focus gradient loss data is subjected to time domain statistics to analyze the influence of parallax changes on eye movement trajectories. Finally, the frequency domain and time domain characteristics are comprehensively converted into a visual fatigue index, and the output is visual perception fatigue data for subsequent design of variable light control mechanisms.

[0140] Step S33: Designing a light variable control mechanism based on the visual perception fatigue data and the visual angle movement light change illusion quantification data and the parallax focus gradient loss data to obtain a light variable control mechanism.

[0141] In an embodiment of the present invention, a variable light control mechanism is designed in combination with visual perception fatigue data, aiming to optimize the 3D image display effect and reduce visual fatigue. First, the light change illusion quantification data is used to analyze the impact pattern of the current light change on visual perception. Subsequently, according to the parallax focus gradient loss data, the main interference direction of the optical parallax on the focus area is identified, and on this basis, the key optical parameters that need to be adjusted are determined. The specific design of the variable light control mechanism includes three core steps: based on the adjustment depth theory, a discrete ray tracing algorithm is used to adjust the light reflection path of different viewing angles in real time to optimize the scattering angle distribution of light in space; using a layered optical adjustment method, the local light intensity of the display screen is compensated for by regional brightness to reduce the burden of high-frequency optical stimulation on the eyes; with the help of dynamic spectral adjustment technology, the wavelength distribution of the color spectrum is analyzed and its local energy distribution is adjusted to make the light in the focus area close to the visual comfort zone.

[0142] Preferably, step S33 includes the following steps:

[0143] Step S331: performing focus drift effect analysis on the perspective movement light change illusion quantification data and the parallax focus gradient loss data based on the visual perception fatigue data to obtain focus drift effect data;

[0144] Step S332: performing color temperature-brightness coupling adjustment analysis on the focus drift effect data to obtain color temperature-brightness coupling adjustment data;

[0145] Step S333: performing local light field rendering deviation correction on the visual angle movement light change illusion quantization data according to the color temperature-brightness coupling adjustment data to obtain local light field rendering deviation correction data;

[0146] Step S334: performing focus space light correction on the focus drift effect data based on the local light field rendering deviation correction data to obtain focus space light correction data;

[0147] Step S335: Designing a light variable control mechanism according to the local light field rendering deviation correction data and the focus space light correction data to obtain a light variable control mechanism.

[0148] In an embodiment of the present invention, the dynamic drift effect of the optical focus is analyzed by inputting visual perception fatigue data, visual angle movement light change illusion quantification data, and parallax focus gradient loss data. First, the focus position trajectory formed by the light in the process of visual angle change is extracted using the illusion quantification data. Subsequently, the temporal and spatial change patterns of the focus drift are multi-dimensionally analyzed in combination with the parallax focus gradient loss data, and the offset and speed of each sampling point are calculated. In a specific implementation, the discrete Fourier transform method is used to project the dynamic change of the light focus into the frequency domain to identify the main frequency components in the drift pattern. Combined with the visual perception fatigue data, the focus drift influence area is segmented and marked, and the contribution of each area to visual fatigue is analyzed. Through interpolation fitting and deviation statistics, the focus drift effect data is generated to provide basic data support for the subsequent color temperature-brightness coupling adjustment. The focus drift effect data is used as input to study the coupling relationship between color temperature and brightness in the optical system. First, according to the light trajectory in the focus drift effect data, the color temperature characteristics of each point are extracted using the color light separation technology based on spectral distribution. The dynamic change trend of brightness is calculated by normalizing the brightness data point by point. The coupling analysis algorithm is used to pair and fit the change patterns of color temperature and brightness, and analyze their synergistic relationship in the focus drift area. In the specific operation, the wavelength-weighted average method is used to quantify the color temperature, and the brightness change is converted into the radiation flux of the corresponding area, and a nonlinear coupling equation group between the two is established. Finally, the color temperature-brightness coupling adjustment data is generated by numerical integration solution to provide the input basis for light field rendering deviation correction. The color temperature-brightness coupling adjustment data is used to correct the rendering deviation in the perspective movement light change illusion quantification data. First, according to the coupling adjustment data, the light field distribution of the inconsistent color temperature and brightness areas is located, and the brightness peak and color temperature mismatch points of these areas are marked. Subsequently, combined with the illusion quantification data, the deviation sources of these light field areas are analyzed, including perspective angle, surface reflection characteristics, etc. In the specific implementation, the rendering correction algorithm based on the optical reflection law is used to reversely trace the light path of the marked area to reconstruct the real light field distribution. For nonlinear deviations, the bilinear interpolation algorithm is applied to smooth the deviation distribution to generate a uniform rendering result. Finally, the local light field rendering deviation correction data is output to provide an accurate optical basis for focal space light correction. The local light field rendering deviation correction data and the focus drift effect data are jointly analyzed to correct and optimize the optical distribution of the focal space. First, according to the light field error distribution in the deviation correction data, the light intensity and direction distribution in the focus drift area are adjusted to form a preliminary correction plan. Subsequently, based on the focus drift effect data, the spherical wavefront tracking method is used to calculate the exact position of each focus in space, and the preliminary correction plan is iteratively adjusted through the optimization algorithm to ensure that the optical focus position is consistent with the actual viewing angle.Finally, the correction effect is verified by ray tracing simulation, and the focus space light correction data is output to provide accurate correction parameters for the design of light variable control mechanism. Combining the local light field rendering deviation correction data and the focus space light correction data, a light variable control mechanism is designed to dynamically adjust the optical characteristics of 3D image display. First, the relationship between light field distribution and focus position is extracted through correction data, and a dynamic control demand model is established. Then, according to the optical control theory, the light field area is partitioned to distinguish high-intensity control areas and low-frequency control areas. In the specific design process, optical polarization control technology is used to adjust the direction of the polarizer to achieve the intensity change of light in a specific area. Combined with dynamic light diffraction technology, the wavefront morphology of the light is regulated to enable it to adapt to the dynamic needs of different viewing angles and focus distributions. Through the real-time feedback mechanism, the control effect is verified and fine-tuned, and finally a light variable control mechanism is generated to provide optimized dynamic optical support for 3D image display.

[0149] Preferably, step S334 includes the following steps:

[0150] Based on the local light field rendering deviation correction data, the focus drift effect data is evaluated for the light field intensity distribution difference between different focus drift directions to obtain the light field intensity distribution difference data;

[0151] Decomposing the light field offset vectors between different focus drift directions on the light field intensity distribution difference data to obtain the light field offset vectors;

[0152] Reconstructing the focus error vector of the focus drift effect data according to the light field offset vector to obtain focus error vector reconstruction data;

[0153] Based on the focus error vector reconstruction data and the local light field rendering deviation correction data, spatial light scattering interference analysis is performed to obtain focus light scattering interference data;

[0154] Focus space light correction is performed on the focus drift effect data according to the focus light scattering interference data and the focus error vector reconstruction data to obtain focus space light correction data.

[0155] In an embodiment of the present invention, local light field rendering deviation correction data and focus drift effect data are used as input to evaluate the difference in light field intensity distribution in different directions of focus drift. First, the light field intensity in each direction is calculated by the optical path integral method, and the light energy density change in a specific angle range during focus drift is extracted. Then, the light field intensity distribution is decomposed into directional components by using a difference analysis algorithm, and the average value and standard deviation of the distribution in each direction are compared to obtain the difference in light field intensity distribution. In a specific implementation, the spatial intensity data is converted into frequency domain information by using Fourier transform, and the characteristic parameters of the high difference direction are extracted by spectrum comparison, and finally the light field intensity distribution difference data is generated to provide accurate input for the subsequent light field offset vector analysis. With the light field intensity distribution difference data as input, the light field offset vector in different focus drift directions is calculated by a vector decomposition algorithm. First, the light field intensity difference is converted into a spatial vector field by using a vector field interpolation technique, and the offset components along different directions are separated. Then, the vector field is segmented and analyzed based on the segmented analytical method to calculate the offset intensity and displacement vector in each direction. The data is double-checked in combination with the standard offset model to correct the vector direction deviation caused by sampling error. Finally, the light field offset vector is generated to provide data support for the reconstruction of the focus error vector. The focus error vector reconstruction takes the light field offset vector and the focus drift effect data as input, and reconstructs the focus error using the multidimensional vector fitting method. First, the focus drift effect data is projected into the direction space of the light field offset vector, and the relationship between the offset vector and the actual focus position is fitted using the least squares method. Then, for each direction offset, the distribution of the error vector in space is calculated using the interpolation method, and the outlier correction is performed on the abnormal points. Finally, the vector superposition method is used to integrate the offsets in each direction to obtain the complete focus error vector distribution, forming the focus error vector reconstruction data, which provides the basic input for the spatial light scattering interference analysis. By introducing the focus error vector reconstruction data and the local light field rendering deviation correction data, the spatial light scattering interference effect is analyzed in detail. First, the optical scattering theory is used to calculate the scattering intensity of light in different media, and the high scattering area is located based on the reconstructed data. Then, the coherence and phase difference between the scattered light are evaluated by the optical interference analysis algorithm, and the key characteristic parameters of the interference fringes are extracted. The interference data is converted into spectral information using fast Fourier transform to evaluate the consistency of light scattering in different regions. Finally, the focal light scattering interference data is generated in combination with the spatial distribution characteristics, laying an optical foundation for focal spatial light correction. The focal light scattering interference data and the focal error vector reconstruction data are used to perform spatial light correction on the focal drift effect data. First, the scattering interference data is combined with the error vector data through the vector field superposition method to calculate the global correction value of the spatial light field. Subsequently, the propagation path of each point in the light field is adjusted based on the light path tracing technology to ensure that the position and direction of the light focus are consistent with the theoretical value. In this process, the point-by-point light intensity adjustment method is used to fine-tune the light field distribution to correct the remaining optical deviations.Finally, the focus spatial light correction data is output to provide accurate correction parameter support for subsequent 3D image display optimization.

[0156] Preferably, the present invention further provides a naked-eye 3D screen, including a naked-eye 3D screen image control processor, for executing the 3D image display method as described above, the naked-eye 3D screen comprising:

[0157] The 3D space geometry model building module is used to obtain the multimedia video playback image; construct the background-foreground subject 3D space geometry model of the multimedia video playback image to obtain the background-foreground subject 3D space geometry model;

[0158] The light change illusion analysis module is used to obtain the user's perspective movement data; according to the user's perspective movement data, the perspective movement light change illusion is quantified for the background-foreground subject 3D space geometric model to obtain the perspective movement light change illusion quantification data; according to the perspective movement light change illusion quantification data, the parallax focus gradient loss is evaluated to obtain the parallax focus gradient loss data;

[0159] A light variable control mechanism design module is used to evaluate visual perception fatigue according to parallax focus gradient loss data to obtain visual perception fatigue data; based on the visual perception fatigue data, the light variable control mechanism is designed for the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain the light variable control mechanism;

[0160] The firmware design module is used to perform logic memory learning on the optically variable control mechanism based on the policy gradient algorithm to obtain the optically variable control logic memory data; perform automated control firmware design based on the optically variable control logic memory data to obtain the optically variable control logic firmware, and embed the optically variable control logic firmware into the naked-eye 3D screen image control processor to execute the 3D image display method.

[0161] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0162] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A 3D image display method, characterized in that: The following steps are involved: Step S1: Acquire multimedia video playback image; Constructing a background-foreground subject 3D space geometry model for a multimedia video playback image to obtain a background-foreground subject 3D space geometry model; Step S2: obtaining user perspective movement data; performing perspective movement light change illusion quantification on the background-foreground subject 3D space geometric model according to the user perspective movement data to obtain perspective movement light change illusion quantification data; performing parallax focus gradient loss evaluation according to the perspective movement light change illusion quantification data to obtain parallax focus gradient loss data; Step S3: evaluating visual perception fatigue according to the parallax focusing gradient loss data to obtain visual perception fatigue data; Based on the visual perception fatigue data, the light change illusion quantification data of perspective movement and the parallax focus gradient loss data, the light variable control mechanism is designed to obtain the light variable control mechanism; Step S4: performing logic memory learning on the light-variable control mechanism based on the policy gradient algorithm to obtain light-variable control logic memory data; According to the optically variable control logic memory data, an automated control firmware design is performed to obtain the optically variable control logic firmware, and the optically variable control logic firmware is embedded into the naked-eye 3D screen image control processor to execute the 3D image display method.

2. The 3D image display method according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Acquire multimedia video playback image; Step S12: performing image edge optimization processing on the multimedia video playback image to obtain video playback image edge optimization data; Step S13: marking the background reference object-foreground subject plane coordinates of the video playback image edge optimization data to obtain the background reference object-foreground subject plane coordinates; Step S14: constructing a background-foreground subject 3D space geometry model for the edge optimization data of the video playback image according to the background reference object-foreground subject plane coordinates to obtain a background-foreground subject 3D space geometry model.

3. The 3D image display method according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: Acquire user perspective movement data; Step S22: Based on the user perspective movement data, the focus area of ​​the background-foreground subject 3D space geometric model between different perspectives is identified to obtain the background-foreground subject perspective focus area data; Step S23: quantifying the perspective movement light change illusion of the background-foreground subject perspective focus area data according to the user perspective movement data to obtain perspective movement light change illusion quantification data; Step S24: performing parallax focus gradient loss evaluation on the background-foreground subject perspective focus area data according to the perspective movement light change illusion quantification data to obtain parallax focus gradient loss data.

4. The 3D image display method according to claim 3, characterized in that: Step S23 includes the following steps: Step S231: simulating and calculating the focus area perspective movement offset of the background-foreground subject perspective focus area data according to the user perspective movement data to obtain the focus area perspective movement offset; Step S232: calculating the horizontal section offset angle of the focus area viewing angle movement offset to obtain horizontal section offset angle data; Step S233: performing a light scattering geometric color wheel illusion evaluation on the background-foreground subject perspective focus area data according to the focus area perspective movement offset and the horizontal section offset angle data to obtain geometric color wheel illusion data; Step S234: quantifying the perspective movement and light change illusion of the background-foreground subject perspective focus area data according to the geometric color wheel illusion data to obtain perspective movement and light change illusion quantification data.

5. The 3D image display method according to claim 4, characterized in that: Step S233 includes the following steps: Performing background-foreground subject geometric morphology analysis on the background-foreground subject perspective focus area data to obtain background geometric morphology data and foreground subject geometric morphology data respectively; Perform geometric surface normal vector interleaving calculation on the background geometric morphology data and the foreground subject geometric morphology data to obtain background geometric morphology normal vector interleaving data and foreground subject geometric morphology normal vector interleaving data respectively; Perform background-foreground subject light scattering spectral interference calculation on the background geometric morphology normal vector interlaced data and the foreground subject geometric morphology normal vector interlaced data to obtain global light scattering spectral interference data; Perform spatial dispersion distribution analysis based on global light scattering spectral interference data to obtain spatial dispersion distribution data; Performing a perspective shift dispersion anisotropy analysis on the spatial dispersion distribution data according to the perspective shift offset of the focus area and the horizontal cross-section offset angle data to obtain perspective shift dispersion anisotropy data; Performing a perspective shift dispersion overlap analysis on the perspective shift dispersion anisotropy data to obtain perspective shift dispersion overlap data; The light scattering geometric color wheel illusion is evaluated based on the viewing angle shift dispersion anisotropy data and the viewing angle shift dispersion overlap data to obtain the geometric color wheel illusion data.

6. The 3D image display method according to claim 3, characterized in that: Step S24 includes the following steps: Step S241: performing parallax depth level analysis on the background-foreground subject perspective focus area data according to the perspective movement light change illusion quantification data to obtain the focus area parallax depth level data; Step S242: calculating the average depth difference between the layers of the focus area parallax depth level data to obtain the average parallax depth difference of the regional level; Step S243: evaluating the focus point imaging clarity loss on the perspective movement light change illusion quantification data according to the regional level parallax depth average difference to obtain the focus point imaging clarity loss data; Step S244: performing parallax focus gradient loss evaluation based on the focus point imaging clarity loss data and the regional level parallax depth average difference to obtain parallax focus gradient loss data.

7. The 3D image display method according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing a focus area spatial deformation analysis on the background-foreground subject 3D space geometric model based on the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain the focus area spatial deformation data; Step S32: evaluating the visual perception fatigue of the focus area spatial deformation data according to the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain visual perception fatigue data; Step S33: Designing a light variable control mechanism based on the visual perception fatigue data and the visual angle movement light change illusion quantification data and the parallax focus gradient loss data to obtain a light variable control mechanism.

8. The 3D image display method according to claim 7, characterized in that: Step S33 includes the following steps: Step S331: performing focus drift effect analysis on the perspective movement light change illusion quantification data and the parallax focus gradient loss data based on the visual perception fatigue data to obtain focus drift effect data; Step S332: performing color temperature-brightness coupling adjustment analysis on the focus drift effect data to obtain color temperature-brightness coupling adjustment data; Step S333: performing local light field rendering deviation correction on the visual angle movement light change illusion quantization data according to the color temperature-brightness coupling adjustment data to obtain local light field rendering deviation correction data; Step S334: performing focus space light correction on the focus drift effect data based on the local light field rendering deviation correction data to obtain focus space light correction data; Step S335: Designing a light variable control mechanism according to the local light field rendering deviation correction data and the focus space light correction data to obtain a light variable control mechanism.

9. The 3D image display method according to claim 7, characterized in that: Step S334 includes the following steps: Based on the local light field rendering deviation correction data, the focus drift effect data is evaluated for the light field intensity distribution difference between different focus drift directions to obtain the light field intensity distribution difference data; Decomposing the light field offset vectors between different focus drift directions on the light field intensity distribution difference data to obtain the light field offset vectors; Reconstructing the focus error vector of the focus drift effect data according to the light field offset vector to obtain focus error vector reconstruction data; Based on the focus error vector reconstruction data and the local light field rendering deviation correction data, spatial light scattering interference analysis is performed to obtain focus light scattering interference data; Focus space light correction is performed on the focus drift effect data according to the focus light scattering interference data and the focus error vector reconstruction data to obtain focus space light correction data.

10. A naked eye 3D screen, characterized in that: A naked-eye 3D screen image control processor is included, which is used to execute the 3D image display method according to claim 1, and the naked-eye 3D screen includes: The 3D space geometry model building module is used to obtain the multimedia video playback image; construct the background-foreground subject 3D space geometry model of the multimedia video playback image to obtain the background-foreground subject 3D space geometry model; The light change illusion analysis module is used to obtain the user's perspective movement data; according to the user's perspective movement data, the perspective movement light change illusion is quantified for the background-foreground subject 3D space geometric model to obtain the perspective movement light change illusion quantification data; according to the perspective movement light change illusion quantification data, the parallax focus gradient loss is evaluated to obtain the parallax focus gradient loss data; A light variable control mechanism design module is used to evaluate visual perception fatigue according to parallax focus gradient loss data to obtain visual perception fatigue data; based on the visual perception fatigue data, the light variable control mechanism is designed for the perspective movement light change illusion quantification data and the parallax focus gradient loss data to obtain the light variable control mechanism; The firmware design module is used to perform logic memory learning on the optically variable control mechanism based on the policy gradient algorithm to obtain the optically variable control logic memory data; perform automated control firmware design based on the optically variable control logic memory data to obtain the optically variable control logic firmware, and embed the optically variable control logic firmware into the naked-eye 3D screen image control processor to execute the 3D image display method.

Images