Real-time ultra-high-definition interactive naked-eye 3D display content synthesis method

Abstract

The invention discloses a real-time ultra-high-definition interactive naked-eye 3D display content synthesis method, which comprises the following steps of S1, firstly, modeling interactive content, such as buildings, mechanical equipment, game scenes and the like, and converting mechanical and building models and the like into a. Fbx format by means of a. Step file format in a unified manner; s2, completing model processing in professional model processing software such as 3DsMax and the like: unifying coordinate axes, and performing surface reduction on the model; and S3, building an interaction scene in the Unity 3D based on the processed model. And S4, for a scene and a model needing to be displayed, generating a virtual parallax camera equivalent to off-axis photography, and controlling the shooting parameters of the parallax camera. And S5, performing hybrid synthesis on the multi-view parallax image based on the texture bitmap, calculating large-scale pixels based on a calculation shader, and shifting a GPU idle time sequence in a Unity 3D script process, and then uniformly entering a rendering assembly line to complete rendering and synthesis of the whole scene. And S6, developing interaction input logic according to different object interaction requirements. The invention provides a calculation three-dimensional display algorithm based on a texture bitmap, ultra-high-definition naked-eye 3D display content can be synthesized in real time under Unity 3D, and a better 3D display effect can be obtained by combining technologies such as equivalent off-axis photography and the like. In addition, the invention also provides a display efficiency optimization method based on a calculation shader, so that the system can run on a lower-configuration host platform, and the problem of higher hardware requirements for real-time ultrahigh-definition naked-eye 3D display content synthesis while the interaction requirements are met is solved.

Description

Technical Field

[0001] The present invention belongs to the field of stereoscopic display, and in particular relates to a method for synthesizing real-time ultra-high-definition interactive naked-eye 3D display content. Background Art

[0002] Glasses-free 3D technology is a technology that enables viewers to obtain stereoscopic vision without relying on visual aids. It has broad development prospects in advertising, 3D games, virtual displays and other fields. The synthesis process and efficiency of three-dimensional images are the main factors affecting the effect and experience of naked-eye 3D display. Existing naked-eye 3D content synthesis methods use either the encoding synthesis of video files to produce multiple images (such as a nine-square grid), or the use of depth map synthesis, which has drawbacks in display quality and interactivity. In addition, when synthesizing ultra-high-definition naked-eye 3D display content in real time, it is necessary to process multiple viewpoint images simultaneously, which places high demands on the hardware of the display synthesis processing platform.

[0003] Therefore, this paper proposes a texture bitmap-based computational 3D display algorithm that can synthesize ultra-high-definition glasses-free 3D display content in real time within Unity3D. Combined with techniques such as equivalent off-axis photography, it achieves superior 3D display effects. Furthermore, a compute shader-based display efficiency optimization method is proposed, enabling the system to run on lower-end host platforms, alleviating the high hardware requirements for real-time ultra-high-definition glasses-free 3D display synthesis while meeting interactive requirements. Summary of the Invention

[0004] To address these issues, the present invention proposes a real-time, ultra-high-definition, interactive, glasses-free 3D display content synthesis method. First, the number of virtual viewpoints is set based on the parameters of the lenticular lens grating glasses-free 3D display. Parallax images are then collected in the Unity3D engine according to the set number of viewpoints. These images are then blended using a shader. Furthermore, compute shaders are used to place a large number of pixel calculations during GPU idle time, optimizing blending efficiency and achieving smooth, high frame rates on lower-end hardware platforms.

[0005] In order to achieve the above object, the technical solution of the present invention is as follows:

[0006] A real-time ultra-high-definition interactive naked-eye 3D content synthesis method comprises the following steps:

[0007] S1: First, model the interactive content, such as buildings, mechanical equipment, game scenes, etc. Mechanical and architectural models need to be uniformly converted using the .step file format and converted to the .fbx format.

[0008] S2: Complete model processing in professional model processing software such as 3DsMax: unify the coordinate axes and reduce the surface of the model.

[0009] S3: Build an interactive scene in Unity3D based on the processed model.

[0010] S4: Generate a virtual parallax camera equivalent to off-axis photography for the scene and model to be displayed, and control the shooting parameters of the parallax camera;

[0011] S5: Multi-view parallax images are mixed and synthesized based on texture bitmaps. Large-scale pixel calculations are moved forward to the idle time of the GPU in the Unity3D script process based on compute shaders. They are then uniformly entered into the rendering pipeline to complete the rendering and synthesis of the entire scene.

[0012] S6: Develop interactive input logic based on different object interaction requirements.

[0013] The relevant parameters of the lenticular lens grating naked-eye 3D display include: grating inclination angle a, grating lines per inch LPI, dots per inch DPI, resolution, and intercept P of a single lenticule.

[0014] The equivalent off-axis photography is achieved by setting the position of the parallax camera and staggering the projection matrix so that the final image is shifted relative to the center of the optical axis, simulating the convergence of the eyes when looking at an object, and providing good positive and negative parallax.

[0015] The shooting parameters of the parallax camera include: focus mode, orientation T, coordinate L, rotation R, rendering level, projection mode, near and far plane distance, FOV, texture resolution w*h, and focal length SW.

[0016] The texture bitmap is a method for locating pixels in a dense viewpoint image. In Unity3D, the pixels in the texture are located through the bitmap.

[0017] The Unity3D script execution process is a fixed script lifecycle management process of the Unity3D engine, which fixes the execution order of the rendering link, physics system, event input, game logic and other links.

[0018] The compute shader is a GPU parallel computing technology that distributes large-scale operations to numerous computing units through thread group scheduling and runs outside the rendering pipeline.

[0019] The aforementioned shifting of large-scale pixel calculations to the GPU idle timing is achieved by using the characteristics of the compute shader being independent of the rendering pipeline and being able to directly interact with the video memory for data. Based on this, data transfer can be completed directly within the GPU.

[0020] The beneficial effects of the present invention are:

[0021] By directly rendering and synthesizing naked-eye 3D display content, the demand for real-time ultra-high-definition interactive naked-eye 3D display content in some fields can be met. Combined with equivalent off-axis photography technology, the displayed 3D images have good quality, strong stereoscopic perception and low crosstalk. At the same time, by optimizing the synthesis method, the ultra-high-definition naked-eye 3D display system can run smoothly on a lower-configuration platform, reducing the hardware cost of ultra-high-definition naked-eye 3D real-time rendering, and achieving good economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] To facilitate understanding of the specific implementation methods of the embodiments of the present invention or the prior art solutions, the following will briefly describe the drawings involved in the relevant embodiments or technical descriptions. It should be noted that the drawings shown below are only some representative embodiments of the present invention. Without departing from the core idea of ​​the present invention, those skilled in the art can also deduce other feasible implementation methods based on the technical solutions disclosed in these drawings.

[0023] Figure 1 Model processing flow chart of this method

[0024] Figure 2 Schematic diagram of the structure of the lenticular lens grating naked-eye 3D display used in this method

[0025] Figure 3 The lenticular grating spectroscopic optical path used in this method

[0026] Figure 4 Schematic diagram of multi-view sub-pixel mapping based on this method

[0027] Figure 5 This is the parallax camera initialization and parameter control logic diagram of this method

[0028] Figure 6 Shader synthesis logic for this method

[0029] Figure 7 Texture bitmap execution flow chart proposed for this method

[0030] Figure 8 The logic diagram of parallel calculation of 3D display by the compute shader of this method

[0031] Figure 9 Figure 1 is an abstract DETAILED DESCRIPTION

[0032] The present invention is described in detail below with reference to the accompanying drawings and embodiments:

[0033] S1: Select the content and objects that need to be interactive, such as buildings, mechanical equipment, etc., and model them.

[0034] S11: Figure 1As shown, the model processing flow of the present invention includes the following steps:

[0035] (1) Export STEP digital model files based on the established mechanical and architectural parts;

[0036] (2) Based on the model exported in step (1), process the model in 3DsMax, including model surface reduction, coordinate axis transformation, etc.;

[0037] (3) Export the FBX model in 3DsMax and then import it into Unity3D to build the subsequent interactive environment.

[0038] S2: Optical derivation and sub-pixel mapping derivation for naked-eye 3D display

[0039] S21: Figure 2 As shown, the structural principle of the present invention to achieve naked eye 3D visual perception is:

[0040] (1) Displaying the synthesized naked-eye 3D video image on a liquid crystal display;

[0041] (2) After the light is split by the lenticular lens covering the display, the image with parallax is accurately projected into the left and right eyes of the person;

[0042] (3) The observed images are synthesized by the human brain to produce stereoscopic vision.

[0043] S22: Further, Figure 3 As shown in the figure, Wp is the sub-pixel width of the LCD panel; D is the distance between the LCD panel and the lenticular lens; p is the intercept of the lenticular lens; L is the distance from the human eye observation plane to the naked eye 3D display; T is the viewpoint spacing, and the number of viewpoints N in this optical path schematic is 4. According to the triangular geometric relationship, it can be deduced that:

[0044]

[0045] The observation range distance can be derived based on the derived formula and the parameters of the cylindrical grating.

[0046] S23: The spectral separation of the cylindrical grating covers the displayed sub-pixels through tiny cylindrical lenses at the pixel level, and separated parallax images can be obtained at a set observation distance. From the above cylindrical grating imaging principle, it can be seen that the cylindrical lens controls the display of viewpoint images by modulating the light emitted by the sub-pixels. The image of each viewpoint needs to collect images of different perspectives of the real object for synthesis. If there are K viewpoints, K perspective parallax images are required, and finally these parallax images need to be accurately mixed according to the cylindrical grating parameters. In naked-eye 3D display, we achieve this by encoding the sub-pixels of the LCD screen. The aspect ratio of the RGB of the liquid crystal sub-pixels suitable for naked-eye 3D display is 3:1. After knowing the intercept, tilt angle and number of viewpoints of the cylindrical grating, the viewpoint image source of each sub-pixel can be deduced by calculation, thereby constructing the mapping of the K viewpoints of each sub-pixel in the entire picture when it is finally displayed. The viewpoint mapping of the pixel is as follows Figure 4 As shown:

[0047] X represents the number of sub-pixels covered by a single lenticular lens, Wp represents the width of each sub-pixel of the liquid crystal display panel, P represents the intercept of a single lenticular lens, Px represents the horizontal component of the intercept, a represents the inclination angle of the lenticular lens grating, and x represents the width of each sub-pixel of the liquid crystal display panel. offset represents the horizontal distance between the sub-pixel and the edge of the lenticular lens, Kn is the viewpoint number corresponding to the sub-pixel, and K is the total number of viewpoints. Assuming that the upper left corner of the lenticular lens and the upper left corner of the LCD panel are completely aligned, it can be deduced that:

[0048] P=P x ·cosa (4)

[0049]

[0050] x offset =(xy·tana)·mod P x (6)

[0051] P x =W p ·X (7)

[0052] On the entire LCD panel, the horizontal and vertical positions of the subpixels are represented by i and j respectively, and the aspect ratio of the subpixels is 3:1, so (x, y) can be expressed as:

[0053] x=i·W p (8)

[0054] y=3·j·W p (9)

[0055] From (4) to (9), we can deduce:

[0056]

[0057] Errors are inevitable when processing and manufacturing lenticular lens naked-eye 3D displays. When the lenticular lens and the LCD panel are assembled, the upper left corner of the lenticular lens and the upper left corner of the LCD panel may not be completely aligned. Assuming that the horizontal assembly error between the upper left corner edge of the lenticular lens and the LCD panel is s, correction is required:

[0058]

[0059] When designing a lenticular lens grating naked-eye 3D display, the number of sub-images X covered by a single lenticular lens, the lenticular lens inclination angle a, and the total number of viewpoints K are all determined. ij If the subscript of is known, the viewpoint from which the sub-pixel value should be taken can be deduced according to the above formula, and the viewpoint source of all sub-pixels on the display can be calculated. The parallax image can be encoded according to this viewpoint mapping to obtain the expected naked-eye 3D display effect.

[0060] S3: Parallax image acquisition in Unity3D

[0061] There are four main stereo camera placement methods: parallel, converging, off-axis, and arc. Each has its own advantages and disadvantages: parallel (parallel optical axis, no keystone distortion, but only negative parallax, small common area); converging (large common area, multiple parallax, but keystone distortion and depth perception issues); off-axis (large common area, no keystone distortion, but requires complex post-processing); and arc (wide-angle coverage, but keystone distortion and vertical parallax). This method implements the equivalent of off-axis photography in Unity3D to capture parallax images.

[0062] There are two main types of scene movement when viewing naked-eye 3D images: movement on the display plane and depth movement along the display axis. Planar motion means that the object moves only on a plane parallel to the screen. Depth motion is defined as the movement of objects as they move closer or further away from the viewer. The effect of objects in the displayed scene entering and exiting the screen is the main perception of stereoscopic vision. When displaying 3D images on a flat stereoscopic screen, binocular parallax is It can be expressed in terms of perspective.

[0063] Let F be the position of the object on the screen, let α be the angle formed by the focal direction when each eye focuses on the screen. When the object is on the screen plane, the binocular angular parallax is 0°, which means there is no parallax between the two retinal images. Let the binocular parallax generated when the gaze point F moves to point A between the screen and the human eye and point B behind the screen be and It can be calculated by the following formula:

[0064]

[0065] PD is the distance between the human eyes. The average pupil distance for adults is between 54 and 74 mm, while for children it is between 43 and 54 mm. In this study, PD was set to the international biological average of 65 mm. Parallax is related to the distance between the two cameras, the camera focal length, and the distance between the object and the camera. The appropriate camera distance for virtual stereo photography is now derived:

[0066] Assume that the distance between an object and a virtual stereo camera in a display scene is z i , the distance between the cameras is B, the focal length of the camera is f, and the minimum horizontal parallax on the imaging plane of the virtual stereo camera is:

[0067]

[0068] Assuming that the parallax at infinity is zero, the parallax range is obtained:

[0069]

[0070] Let the magnification of the final image displayed on the screen compared to the image captured by the virtual stereo camera be m, and formula (4) becomes:

[0071]

[0072] Assume that the maximum positive disparity and the maximum negative disparity that the human eye can observe normally for stereo images are d + and d - , then the parallax within the normal display range should satisfy:

[0073] d - <d<d + (6)

[0074] From this we have:

[0075]

[0076] The relationship between camera spacing, scene data, and camera parameters can be derived:

[0077]

[0078] According to formula (8), when the virtual stereo camera collects the parallax picture, the camera placement distance is set according to the captured scene, virtual camera parameters and appropriate parallax range.

[0079] S21: In Unity3D, stereoscopic shooting with a virtual camera involves a key coordinate transformation process: first, the object is transformed from the local coordinate system to the world coordinate system, then transformed to the camera-centered observation space, then mapped to the viewing frustum through the projection matrix for clipping, and finally converted to screen pixel coordinates. The specific implementation is divided into three stages:

[0080] (1) View transformation: The camera’s Transform component (Position / Rotation and other parameters) is used to precisely control the camera’s position and generate a camera sequence based on the parallax design principle.

[0081] (2) Projection transformation: Use the perspective mode to simulate the real perspective, define the view volume through parameters such as FOV and ClippingPlanes, and ensure the correct projection of 3D objects.

[0082] (3) Viewport transformation: The size and position of the rendering window are controlled by the Viewport Rect parameter to adapt to the screen or VR device (a customized projection matrix is ​​required to avoid distortion).

[0083] Off-axis photography requires adjusting the focus and asymmetric viewing angle of the off-axis stereo camera based on the distance to the object being viewed, which requires extensive calculations. Equivalent off-axis photography can be achieved using an oblique projection matrix: by constructing an oblique projection matrix with asymmetric projection properties, it effectively simulates the perspective distortion caused by the spatial offset between the camera position and the object center, referencing how the human eye actually observes an object.

[0084] In off-axis photography, there is a positional offset between the camera position and the center of the object, which causes the perspective projection matrix to change. The physical displacement of off-axis photography is converted into a mathematical transformation of the projection matrix parameters. By adjusting the shear parameters of the projection matrix, an equivalent off-axis photography effect can be achieved. The following analyzes the implementation of equivalent off-axis photography in Unity3D.

[0085] The mathematical form of the standard perspective projection matrix P is as follows:

[0086]

[0087] Where asp is the camera aspect ratio, far is the camera far clipping plane distance, near is the camera near clipping plane distance, v F is the vertical field of view angle, which is directly given by the camera's FOV and is expressed in radians. F is the horizontal field of view angle, which is calculated by the vertical field of view angle and the camera aspect ratio:

[0088]

[0089] To obtain the projection shear parameters that modify the standard perspective projection matrix P, it is necessary to standardize the offset in the world space and define the standardized horizontal offset as X offset , the vertical offset is Y offset , and the horizontal offset in world space is H offset , the vertical offset is V offset , the distance from the camera to the object is D f , then the following calculation formula is available:

[0090]

[0091] After modifying the shear parameters, the standard perspective projection matrix P becomes a custom projection matrix P ob :

[0092]

[0093] Write a script in Unity3D to automatically implement custom projection calculation and mounting application for each parallax camera, and finally obtain the equivalent off-axis photography parallax image in the Render Texture.

[0094] S22: After completing the setting of the equivalent off-axis photography parallax camera, adjust the parallax camera in the program, such as Figure 5 shown.

[0095] S3: Perform parallax image blending based on texture bitmap method

[0096] The process of obtaining viewpoint parallax images in this method is carried out in parallel. In the pixel arrangement design of the display end, the sub-pixel array follows a predefined mapping rule that is independent of the display content. That is, the arrangement pattern of all sub-pixels in the displayed image is independent of the specific displayed content, and the synthesis of pixels at fixed positions in each frame is fixed.

[0097] Based on this, multi-viewpoint image synthesis can be completed through real-time analysis of disparity information: taking the dual-viewpoint mode as an example, in the final displayed three-dimensional image, the pixels that need to be displayed on the left and right disparity images can be calculated. For the cylindrical lens grating and liquid crystal display panel with specific parameters that need to be coupled, the sub-pixels of different viewpoints are spatially rearranged according to a specific phase offset rule. After the optical splitting effect of the grating, the rearranged composite image can achieve physical separation of the left and right eye disparity light paths, and finally form a three-dimensional image with depth perception on the retina of the human eye.

[0098] In the multi-view mapping algorithm discussed in this article, the horizontal coordinate increases corresponding to the arrangement of RGB subpixels, while the vertical coordinate remains constant at the pixel level. This design stems from the physical properties of LCD backplanes, where each pixel unit consists of three horizontally arranged R, G, and B subpixels. In the programming implementation of dense parallax image blending, the parallax images are processed through a pipeline and ultimately assigned pixel-level values ​​by the fragment shader. In this case, the calculation of each pixel relies on a unique UV coordinate positioning mechanism.

[0099] To achieve precise sub-pixel control, this paper defines three-dimensional coordinates (i, j, n), where i represents the horizontal pixel coordinate (corresponding to the physical column number of the display panel), j represents the vertical pixel coordinate (corresponding to the physical row number of the display panel), and n∈{0,1,2} represents the sub-pixel ordinal number (corresponding to the R, G, and B components, respectively). Taking a 3840×2160 resolution experimental device as an example, its sub-pixel matrix can be decomposed into a horizontally expanded 11520×2160 array. Each logical pixel is horizontally expanded into three independent sub-pixel units. The mathematical relationship between sub-pixel position and multi-viewpoint mapping can be derived:

[0100]

[0101] Under the UV coordinate normalization feature in Unity3D, sub-pixel parallax control is achieved through 3D coordinate expansion, effectively handling the viewpoint offset problem of different color channels in multi-view image synthesis. According to this calculation method, the texture bitmap of each Render Texture is calculated, and finally the output is superimposed and the color value is returned. The above logic is written in the shader and then mounted on the rendering material to obtain the output of the final mixed parallax image. The algorithm flow of implementing texture bitmap calculation and mixing logic through shaders is as follows Figure 6 shown.

[0102] After the shader code is implemented, to obtain stable and interactive naked-eye 3D display content, it is necessary to write a script program coupled with the parallax camera and a control script program in Unity3D. The rendering synthesis process of the entire life cycle is briefly as follows: Figure 7 shown.

[0103] S4: Computational efficiency optimization based on compute shaders

[0104] In Unity3D, a Render Texture is a dynamically generated texture type. Its core principle is to output the camera's rendering results directly to the texture rather than the screen, thereby enabling real-time capture and reuse of scene content. Render Texture is essentially achieved by associating the GPU's FrameBuffer (frame buffer object) with a texture object. The FrameBuffer is the buffer on the GPU that stores rendering results (such as color, depth, and other data). Render Texture directly binds this buffer's data to the texture through APIs (such as FramebufferTexture2D), avoiding the traditional CPU-GPU data transfer process. Based on this, parallax image data can be manipulated in video memory.

[0105] Compute shader is a parallel computing technology that can be independent of the rendering pipeline. It can perform calculations when the GPU is idle and execute some GPGPU algorithms. In Unity3D, script events are executed in a certain order. The rendering of each frame (Scene rendering) is performed after the physics system (Physics), mouse event input (Input events), and game logic (Game logic). That is, after completing all the related processing of scene objects, they enter the rendering pipeline. This means that there is a resource utilization window for the GPU after completing the previous calculation tasks. During this window, the subsequent large-scale mask calculations are pre-placed, and the data is structured and stored in the video memory. After the render texture (Render Texture) is loaded into the video memory, the GPU atomic operation is directly called to perform pixel-level parallel color mixing. The logic of the whole process is as follows Figure 8 shown.

[0106] The core of this method's performance optimization lies in the memory residency strategy of the heterogeneous computing architecture. Data flow analysis shows that during the entire processing cycle, only a single structured data transfer of mask parameters occurs between the CPU and GPU. Subsequent computational tasks such as mask generation, multi-level texture sampling, and pixel synthesis are all completed in a closed-loop on the GPU. Compared to the traditional CPU-GPU frequent data exchange model, this architecture fully utilizes the advantages of memory bandwidth (GDDR6X memory bandwidth is several times the PCIe 4.0 bus bandwidth) and GPU parallel processing speed. Based on the characteristics of the compute shader, complex calculations are distributed to each thread, greatly accelerating the calculation speed.

[0107] Example: Based on two computational 3D display methods, rendering experiments were carried out in the same scene: Unity3D's Profiler and Profile Analyzer were used to record and analyze the operation status. The experiments were carried out on two test platforms. Their hardware parameters are shown in Table 1, and the software environment is Unity3D2022.3.17f1c1, DirectX11, and UniversalRP14.0.9.

[0108] Table 1 Test platform hardware parameters

[0109]

[0110]

[0111] The experimental data after running 200 frames in the same 4K scene is as follows:

[0112] The average frame rate of test platform 1 before optimization was 58.30 FPS, and after optimization, it reached 82.98 FPS, a relative performance gain of 42.33%. The overall frame rate was relatively stable, with no prolonged periods of excessive time spent on a single frame.

[0113] The average frame rate of test platform 2 before optimization was 136.79 FPS, and after optimization, it reached 156.25 FPS, a relative performance gain of 14.2%. The overall frame rate was relatively stable, with no extended periods of excessive time spent on a single frame. Experimental data demonstrated a significant increase in computational efficiency after architectural optimization, validating the engineering effectiveness of this optimization solution in a real-world operating environment.

[0114] Analysis of experimental data reveals that optimization achieves higher frame rate gains on lower-end display devices, primarily due to fundamental differences in resource allocation, computing power, and architectural characteristics between low-end and high-end hardware. Low-end GPUs have fewer parallel computing units, fewer registers, and less cache capacity, making them more susceptible to computing resource bottlenecks. Furthermore, there's a gap in video memory bandwidth, resulting in slower parallel processing. The optimization method implemented in this paper achieves significant frame rate improvements on low-end devices, demonstrating its ability to overcome hardware bottlenecks and highlighting the core role of the "barrel effect" in optimization: optimizing the resources with the smallest gap yields the greatest benefits.

[0115] This is of great significance to the cost of system implementation, as it allows the required host configuration under the same business load to be reduced to an 8th-generation i5 processor and 8GB of memory, and the graphics card is equivalent to the performance level of GeForce GTX 1050. This means that the system can be run on a host with lower configuration, greatly reducing hardware costs and achieving important economic benefits.

[0116] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and it should be understood that the scope of protection of the present invention is not limited to such specific descriptions and embodiments. Those skilled in the art can make various other specific variations and combinations based on the technical teachings disclosed in the present invention without departing from the essence of the present invention, and such variations and combinations are still within the scope of protection of the present invention.

Claims

1. A method for synthesizing real-time ultra-high-definition interactive naked-eye 3D display content, comprising the following steps: S1: First, model the content that needs to be interactively displayed, such as buildings, mechanical equipment, game scenes, etc. The mechanical and architectural models need to be uniformly converted using the .step file format and converted to the .fbx format; S2: Complete the model processing in professional model processing software such as 3DsMax: unify the coordinate axis and reduce the surface of the model; S3: Build an interactive scene in Unity3D based on the processed model; S4: Generate a virtual parallax camera equivalent to off-axis photography for the scene and model to be displayed, and control the shooting parameters of the parallax camera; S5: Multi-viewpoint parallax images are mixed and synthesized based on texture bitmaps. Large-scale pixel calculations are moved forward to the idle GPU timing in the Unity3D script process based on compute shaders. They are then uniformly entered into the rendering pipeline to complete the rendering and synthesis of the entire interactive scene. S6: Develop interactive input logic based on different object interactive display requirements.

2. A method for synthesizing real-time ultra-high-definition interactive naked-eye 3D display content according to claim 1, characterized in that The lenticular lens grating naked-eye 3D display is used to achieve the naked-eye 3D display effect by re-arranging the pixels of the display screen and splitting the light through the grating.

3. The method for synthesizing real-time ultra-high-definition interactive naked-eye 3D display content according to claim 1, characterized in that Based on the Unity3D engine, the synthesis of ultra-high-definition real-time interactive naked-eye 3D display content is realized, and texture bitmaps are used to realize the synthesis of parallax images.

4. The method for synthesizing real-time ultra-high-definition interactive naked-eye 3D display content according to claim 1, characterized in that Compute shaders are used to accelerate dense viewpoint calculations, improving computational efficiency by utilizing GPU idle timing throughout the script lifecycle.

Images