Control method of naked eye 3D display device

Abstract

The invention relates to the technical field of naked-eye 3D display, and discloses a control method of a naked-eye 3D display device, which comprises the following steps of: S1, detecting a viewpoint of a user and acquiring parameters, acquiring the position, the viewing angle and the viewing distance of the eyes of the user through a sensor, and calculating a theoretical parallax value based on the acquired parameters; and S2, dynamic optimal parallax matching calculation: on the basis of obtaining the eye position, the viewing angle and the viewing distance of the user, combining depth information of the display content, and calculating optimal parallax distribution in the current environment. A dynamic optimal parallax matching technical scheme is adopted, user eye movement tracking information and display content depth information are combined, parallax distribution is optimized in real time, it is ensured that a 3D picture is always in an optimal watching state, compared with an existing naked eye 3D display scheme with fixed parallax parameters, the method can adapt to watching angles, eye distances and distances of different users, and the user experience is improved. The problem that a fixed parallax scheme is poor in adaptability is solved, the watching comfort is greatly improved, and visual fatigue is reduced.

Description

Technical Field

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

[0002] Glasses-free 3D display technology adjusts the image information received by the left and right eyes, allowing viewers to experience a stereoscopic visual effect without the need for specialized glasses. Current technologies primarily rely on methods such as parallax barriers, lenticular lens arrays, or multi-viewpoint light field displays to reconstruct light from different angles, enabling the human eye to perceive 3D images. However, existing technical solutions still have many limitations in practical applications, affecting the 3D display quality and user experience.

[0003] First, fixed parallax solutions lack adaptability. Traditional naked-eye 3D displays typically use preset parallax parameters, but users' eye spacing, viewing distance, and viewing angle vary, leading to varying user experiences when viewing the same 3D image. For example, some users experience a lack of depth perception, while others may experience visual fatigue due to parallax mismatch. Existing technologies lack dynamic adjustment mechanisms, making it difficult to adapt to individual viewing needs.

[0004] Secondly, the lens array has limited control capabilities. Current cylindrical lenses or MEMS microlenses can adjust the light field to a certain extent, but the focus adjustment method is usually discrete, and continuous and smooth focus changes are impossible. For example, some MEMS lens-based solutions can change the focus through mechanical drive, but the adjustment speed is slow and is prone to jitter and noise due to the mechanical structure. Due to the limited focus adjustment accuracy, the picture is prone to parallax jumps, artifacts, and ghosting, which affect the 3D viewing experience.

[0005] Furthermore, computational efficiency is low and real-time performance is insufficient. Glasses-free 3D requires complex calculations of the light field to ensure accurate matching of images from all perspectives. Existing CPU computing solutions are computationally intensive when processing parallax optimization, light field manipulation, and 3D rendering, resulting in slow system response. Parallax update delays are particularly common in dynamic scenes, impacting smoothness and immersion. Some GPU acceleration solutions can improve computing speed to a certain extent, but they still struggle to meet the demands of high-precision, low-latency, real-time 3D processing.

[0006] Therefore, those skilled in the art provide a control method for a naked-eye 3D display device to solve the above-mentioned problems. Summary of the Invention

[0007] In view of the shortcomings of the prior art, the present invention provides a control method for a naked-eye 3D display device to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: A control method for a naked-eye 3D display device, comprising:

[0009] Step S1: user viewpoint detection and parameter acquisition, using sensors to collect the user's eye position, viewing angle, and viewing distance, and calculating theoretical parallax values based on the acquired parameters;

[0010] Step S2: Dynamic optimal disparity matching calculation, based on the user's eye position, viewing angle and viewing distance, combined with the depth information of the displayed content, calculates the optimal disparity distribution in the current environment;

[0011] Step S3: light field control. After calculating the optimal parallax distribution, the light field distribution of the display screen is adjusted according to the distribution.

[0012] Step S4: driving the variable focus liquid crystal lens array, controlling the refractive index of the liquid crystal lens on the basis of adjusting the light field distribution of the display screen, and realizing dynamic focal length adjustment;

[0013] Step S5, adaptively adjust the display content. After the liquid crystal lens array completes the dynamic focal length adjustment, the display content is optimized and rendered to match the current optimal parallax distribution. Then, the system returns to step S1 to continue user viewpoint detection and parameter acquisition, forming a real-time closed-loop control.

[0014] Preferably, in step S1, user viewpoint detection and parameter acquisition further include:

[0015] Step 1.1, user eye tracking: Use the eye tracking device to obtain the user's eye position and gaze direction in real time, and set the eye tracking device to output the current eye position vector and gaze direction vector;

[0016] Eye position: P eye =(x eye ,y eye ,z eye ),

[0017] Among them, P eye is the position coordinate of the user's eyes, x eye 、y eye 、z eye Represents the coordinates of the user's eyes in three-dimensional space;

[0018] Sight direction: D view =(d x ,d y ,d z ),

[0019] Among them, D view is the direction vector of the user's line of sight, d x d yd z Represents the components of the sight direction on the x, y, and z coordinate axes;

[0020] Step 1.2, calculate the viewing distance and pupil distance: Calculate the viewing distance D by using the eye position and line of sight obtained by the eye tracking device and combining it with the known geometric parameters of the display. obs and pupil distance P interpupillary ;

[0021]

[0022] Among them, (x screen ,y screen ,z screen ) is the spatial coordinate of a point on the display screen, D obs is the observation distance;

[0023] P interpupillary =|x left -x right |,

[0024] Among them, P interpupillary is the pupil distance, x left 、x right are the position coordinates of the left eye and the right eye;

[0025] Step 1.3, generate theoretical disparity distribution: calculate theoretical disparity δ using the Horopter curve theory The Horopter curve describes the trajectory of the point with the smallest parallax of the human eye. The formula is as follows:

[0026]

[0027] Among them, δ theory is the theoretical parallax, b z is the depth of the object;

[0028] Step 1.4, obtain the display content depth map: used to obtain the depth information of the current 3D display content, set the depth map of the display content to D depth (x, y) represents the depth value of each pixel in the display content;

[0029] Depth Map D depth (x,y):D depth (x, y) is the pixel depth at position (x, y) in the display coordinate system, which is extracted from the 3D model using a computer vision algorithm and represents the relative distance of objects in the displayed content.

[0030] Preferably, in step S2, the dynamic optimal disparity matching calculation further includes:

[0031] Step 2.1, set the objective function and solve the optimal disparity distribution: To optimize the disparity matching effect, set the objective function. The calculated optimal disparity distribution can meet the requirements of minimizing the visual fusion error, image content constraint error, temporal continuity error, and light field optimization error. The objective function is defined as follows:

[0032] E(δ)=w1E fusion (δ)+w2E content (δ)+w3E temporal (δ)+w4E lightfield (δ),

[0033] Among them, E(δ) is the total optimization objective function of the optimal disparity distribution, E fusion (δ) is the visual fusion error, E content (δ) is the image content constraint error, E temporal (δ) is the time continuity error, E lightfield (δ) is the light field optimization error, w1, w2, w3, w4 are the weight coefficients of each error term;

[0034] Step 2.2, establish the parallax optimization equation: After the objective function is set, build a parallax optimization solution model based on the Euler-Lagrange equation. The expression of the Euler-Lagrange equation is as follows:

[0035] Among them, δ is the currently calculated disparity distribution, x and y are the spatial coordinates on the display screen, t is the time, and δ x , δ y is the partial derivative of the disparity distribution in the spatial direction, δ t is the partial derivative of the disparity distribution in time, and d is a constant term.

[0036] Preferably, in step S2, the dynamic optimal disparity matching calculation further includes:

[0037] Step 2.3: Solve using numerical optimization methods: Since the analytical solution of the Euler-Lagrange equation is relatively complex, a numerical optimization method is used to calculate to obtain the optimal disparity distribution. Specific methods include the gradient descent method and the conjugate gradient method. The calculation formula of the gradient descent method is as follows:

[0038] Among them, δ (n) is the disparity distribution of the nth iteration, η is the learning rate, is the objective function gradient under the current disparity distribution, δ (n+1) is the disparity distribution of the nth iteration;

[0039] If the conjugate gradient method is used, the update formula is as follows:

[0040] δ (n+1) =δ (n) +αp (n) ,

[0041]

[0042] Among them, p (n) is the conjugate direction of the nth iteration, α is the step size factor, β is the conjugate coefficient, is the objective function gradient under the parallax distribution at the next moment,

[0043] p (n+1) is the conjugate direction of the n+1th iteration;

[0044] After completing the numerical optimization calculation, the final optimal disparity distribution is obtained, and the optimal disparity distribution will be used for light field control in step S3.

[0045] Preferably, in step S3, light field control further includes:

[0046] Step 3.1, calculate the target light field distribution: After obtaining the optimal disparity distribution, the next step is to calculate the target light field distribution based on the disparity optimization results. The target light field distribution L target (x,y) is the optimal disparity distribution δ calculated based on optimal (x,y) and display content depth map D depth The relationship between (x, y) is obtained;

[0047] Target light field distribution L target (x,y) is given by:

[0048]

[0049] Among them, L target (x, y) is the target light field distribution, D depth (x, y) is the depth value of the display content at position (x, y), δ optimal (x,y) is the optimal disparity distribution, D obs (x, y) is the viewing distance between the user and the display screen;

[0050] Step 3.2, calculate the light field error: After calculating the target light field distribution, evaluate the deviation between the current display light field and the target light field, and define the light field error E f (D) is used to measure the difference between the current light field and the target light field. The light field error formula is as follows:

[0051] E f (D)=∑ x,y (L current (x,y)-Ltarget (x,y)) 2 ,

[0052] Among them, E f (D) is the light field error, L current is the light field distribution of the current display, L target (x,y) is the target light field distribution;

[0053] Step 3.3, adjust the light field generation algorithm: Based on the calculated light field error, the next step is to adjust the light field generation algorithm so that the light field distribution tends to the target light field distribution and achieve parallax matching. The light field adjustment formula is as follows:

[0054]

[0055] Among them, L adjusted (x, y) is the light field distribution after adjustment, L current (x, y) is the current light field distribution, L target (x, y) is the target light field distribution, and α is the adjustment coefficient.

[0056] Preferably, in step S4, driving the variable focus liquid crystal lens array further includes:

[0057] Step 4.1, calculate the refractive index of the liquid crystal lens: After the light field control is completed, calculate the refractive index of the liquid crystal lens at different voltages. The refractive index n(V) of the liquid crystal lens is controlled by the electric field. The calculation formula is as follows:

[0058]

[0059] Where n(V) is the refractive index of the liquid crystal lens under the action of voltage V, n o is the normal refractive index of the liquid crystal material, n e is the anomalous refractive index of the liquid crystal material, V is the voltage applied to the liquid crystal lens, V t is the characteristic voltage that the liquid crystal molecules respond to;

[0060] Step 4.2, calculate the focal length of the liquid crystal lens: After calculating the refractive index of the liquid crystal lens, calculate the focal length f(V) based on the refractive index to determine the imaging characteristics of the liquid crystal lens at different voltages. The formula for calculating the focal length of the liquid crystal lens is as follows:

[0061] Where f(V) is the focal length of the liquid crystal lens under the action of voltage V, r is the radius of curvature of the liquid crystal lens, and n(V) is the refractive index of the liquid crystal lens calculated in step 4.1;

[0062] Step 4.3: Drive the liquid crystal lens array. After calculating the refractive index and focal length of the liquid crystal lens, the next step is to control the voltage of the liquid crystal lens array so that the lens focal length can match the optimal parallax distribution in real time. The driving control equation of the liquid crystal lens is as follows:

[0063]

[0064] Among them, V control (x, y) is the control voltage of the lens array at position (x, y), V min 、V max is the minimum and maximum operating voltage of the lens array, f target (x,y) is the target focal length, f min 、f max are the minimum and maximum focal lengths of the liquid crystal lens.

[0065] Preferably, in step S5, the adaptive adjustment of the display content further includes:

[0066] Step 5.1, optimize 3D image rendering parameters: After the focal length of the liquid crystal lens is adjusted, optimize the rendering parameters of the 3D image so that the 3D image can match the optimal parallax distribution δ optimal (x,y) and the focal length of the liquid crystal lens f(V), the calculation method is as follows:

[0067] Depth of Field Adjustment

[0068]

[0069] in, is the optimized depth of field value, D depth (x, y) is the depth map of the original display content, δ optimal (x,y) is the optimal disparity distribution, δ theory (x, y) is the theoretical parallax distribution, λ is the depth compensation coefficient;

[0070] Hierarchical Hybrid Optimization blend (x,y):

[0071] L blend (x,y)=w f ·L foreground (x,y)+w b ·L background (x,y),

[0072] Among them, L blend (x, y) is the brightness value of the mixed image, L foreground (x,y) and L background (x,y) represents the brightness distribution of the foreground and background, w f 、wb is the weight coefficient of foreground and background;

[0073] Focal length compensation f compensated (x,y):

[0074]

[0075] Among them, f compensated (x,y) is the focal length after compensation, f(V) is the focal length of the liquid crystal lens, D obs (x,y) is the observation distance, and μ is the focal length compensation factor.

[0076] Preferably, in step S5, the adaptive adjustment of the display content further includes:

[0077] Step 5.2, adjust the inter-frame interpolation algorithm: After optimizing the 3D image rendering parameters, adjust the inter-frame interpolation algorithm to improve the smoothness of dynamic 3D display and reduce the visual delay caused by parallax adjustment and light field changes;

[0078] Inter-frame interpolation calculation:

[0079] I new (x, y, t) = (1-β)·I prev (x,y,t)+β·I next (x,y,t),

[0080] Among them, I new (x, y, t) is the interpolated frame image, I prev (x,y,t) and I next (x, y, t) represents the image data of the previous frame and the next frame, and β is the interpolation weight;

[0081] Step 5.3: Use GPU parallel computing to accelerate 3D rendering: After completing image rendering parameter optimization and inter-frame interpolation adjustment, in order to improve the real-time performance of the system, use GPU parallel computing to accelerate the 3D rendering process;

[0082] Parallel computing framework:

[0083] Among them, T render is the rendering time after GPU parallel computing, T CPU N is the rendering time of CPU serial calculation, GPU Calculate the number of cores for the GPU.

[0084] A terminal device includes a memory, a processor and a display device, wherein the processor is used to execute a control method for a naked-eye 3D display device.

[0085] A storage medium stores computer-readable instructions. When the instructions are executed by a terminal device, the terminal device executes a control method for a naked-eye 3D display device.

[0086] The present invention provides a control method for a naked-eye 3D display device. It has the following beneficial effects:

[0087] 1. This invention adopts a dynamic optimal parallax matching technology solution, combining user eye tracking information with display content depth information to optimize the parallax distribution in real time, ensuring that the 3D image is always in the best viewing state. Compared with existing naked-eye 3D display solutions with fixed parallax parameters, this invention can adapt to different users' viewing angles, eye distances, and distances, solving the poor adaptability of fixed parallax solutions, significantly improving viewing comfort and reducing visual fatigue.

[0088] 2. The present invention dynamically adjusts the focal length through a variable-focus liquid crystal lens array, precisely controlling the light field distribution and ensuring smooth and natural parallax matching of the displayed content. Compared to traditional MEMS lenses or fixed-focal-length optical lens solutions, the liquid crystal lens control system of the present invention can achieve millisecond-level dynamic adjustment and continuously adjustable focal length, thus avoiding the parallax jump problem caused by discrete adjustment and reducing artifacts and visual impact.

[0089] 3. The present invention uses GPU parallel computing to optimize the parallax matching solution process, accelerating 3D rendering and light field control. Compared with traditional CPU computing optimization solutions, the GPU computing architecture of the present invention can simultaneously process large amounts of pixel-level data, achieve millisecond-level parallax updates, avoid the dizziness caused by low frame rate and delay, and improve the smoothness and immersion of 3D interaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 Flowchart of the present invention. DETAILED DESCRIPTION

[0091] To help those skilled in the art understand the present invention, the following will provide a clear and complete description of the technical solutions in the embodiments of the present invention, in conjunction with the accompanying drawings. Obviously, the described embodiments are only partial embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0092] The present invention is described in detail below with reference to the accompanying drawings:

[0093] Example:

[0094] Please see the attached Figure 1 An embodiment of the present invention provides a method for controlling a naked-eye 3D display device, comprising:

[0095] Step S1: user viewpoint detection and parameter acquisition, using sensors to collect the user's eye position, viewing angle, and viewing distance, and calculating theoretical parallax values based on the acquired parameters;

[0096] Step 1.1, user eye tracking: Use the eye tracking device to obtain the user's eye position and gaze direction in real time, and set the eye tracking device to output the current eye position vector and gaze direction vector;

[0097] Eye position: P eye =(x eye ,y eye ,z eye ),

[0098] Among them, P eye is the position coordinate of the user's eyes, x eye 、y eye 、z eye Represents the coordinates of the user's eyes in three-dimensional space;

[0099] Sight direction: D view =(d x ,d y ,d z ),

[0100] Among them, D view is the direction vector of the user's line of sight, d x d y d z Represents the components of the sight direction on the x, y, and z coordinate axes;

[0101] Step 1.2, calculate the viewing distance and pupil distance: Calculate the viewing distance D by using the eye position and line of sight obtained by the eye tracking device and combining it with the known geometric parameters of the display. obs and pupil distance P interpupillary ;

[0102]

[0103] Among them, (x screen ,y screen ,z screen ) is the spatial coordinate of a point on the display screen, D obs is the observation distance;

[0104] P interpupillary =|x left -x right |,

[0105] Among them, P interpupillary is the pupil distance, x left 、x right are the position coordinates of the left eye and the right eye;

[0106] Step 1.3, generate theoretical disparity distribution: calculate theoretical disparity δ using the Horopter curve theory The Horopter curve describes the trajectory of the point with the smallest parallax of the human eye. The formula is as follows:

[0107]

[0108] Among them, δ theory is the theoretical parallax, b z is the depth of the object;

[0109] Step 1.4, obtain the display content depth map: used to obtain the depth information of the current 3D display content, set the depth map of the display content to D depth (x, y) represents the depth value of each pixel in the display content;

[0110] Depth Map D depth (x,y):D depth (x, y) is the pixel depth at position (x, y) in the display coordinate system, which is extracted from the 3D model using a computer vision algorithm and represents the relative distance of objects in the displayed content;

[0111] Step S2: Dynamic optimal disparity matching calculation, based on the user's eye position, viewing angle and viewing distance, combined with the depth information of the displayed content, calculates the optimal disparity distribution in the current environment;

[0112] In step S2, the dynamic optimal disparity matching calculation further includes:

[0113] Step 2.1, set the objective function and solve the optimal disparity distribution: To optimize the disparity matching effect, set the objective function. The calculated optimal disparity distribution can meet the requirements of minimizing the visual fusion error, image content constraint error, temporal continuity error, and light field optimization error. The objective function is defined as follows:

[0114] E(δ)=w1E fusion (δ)+w2E content (δ)+w3E temporal (δ)+w4E lightfield (δ),

[0115] Among them, E(δ) is the total optimization objective function of the optimal disparity distribution, E fusion (δ) is the visual fusion error, E content (δ) is the image content constraint error, E temporal (δ) is the time continuity error, E lightfield (δ) is the light field optimization error, w1, w2, w3, w4 are the weight coefficients of each error term;

[0116] Step 2.2, establish the parallax optimization equation: After the objective function is set, build a parallax optimization solution model based on the Euler-Lagrange equation. The expression of the Euler-Lagrange equation is as follows:

[0117] Among them, δ is the currently calculated disparity distribution, x and y are the spatial coordinates on the display screen, t is the time, and δ x , δ y is the partial derivative of the disparity distribution in the spatial direction, δ t is the partial derivative of the parallax distribution in time, and d is a constant term;

[0118] Step 2.3: Solve using numerical optimization methods: Since the analytical solution of the Euler-Lagrange equation is relatively complex, a numerical optimization method is used to calculate to obtain the optimal disparity distribution. Specific methods include the gradient descent method and the conjugate gradient method. The calculation formula of the gradient descent method is as follows:

[0119]

[0120] Among them, δ (n) is the disparity distribution of the nth iteration, η is the learning rate, is the objective function gradient under the current disparity distribution, δ (n+1) is the disparity distribution of the nth iteration;

[0121] If the conjugate gradient method is used, the update formula is as follows:

[0122] δ (n+1) =δ (n) +αp (n) ,

[0123]

[0124] Among them, p (n) is the conjugate direction of the nth iteration, α is the step size factor, β is the conjugate coefficient, is the objective function gradient under the parallax distribution at the next moment,

[0125] p (n+1) is the conjugate direction of the n+1th iteration;

[0126] After completing the numerical optimization calculation, the final optimal disparity distribution is obtained, which will be used for light field control in step S3;

[0127] Step S3: light field control. After calculating the optimal parallax distribution, the light field distribution of the display screen is adjusted according to the distribution.

[0128] Step 3.1, calculate the target light field distribution: After obtaining the optimal disparity distribution, the next step is to calculate the target light field distribution based on the disparity optimization results. The target light field distribution L target (x,y) is the optimal disparity distribution δ calculated based on optimal (x,y) and display content depth map D depth The relationship between (x, y) is obtained;

[0129] Target light field distribution L target (x,y) is given by:

[0130]

[0131] Among them, L target (x, y) is the target light field distribution, D depth (x, y) is the depth value of the display content at position (x, y), δ optimal (x,y) is the optimal disparity distribution, D obs (x, y) is the viewing distance between the user and the display screen;

[0132] Step 3.2, calculate the light field error: After calculating the target light field distribution, evaluate the deviation between the current display light field and the target light field, and define the light field error E f (D) is used to measure the difference between the current light field and the target light field. The light field error formula is as follows:

[0133] E f (D)=∑ x,y (L current (x,y)-L target (x,y)) 2 ,

[0134] Among them, E f (D) is the light field error, L current is the light field distribution of the current display, L target (x,y) is the target light field distribution;

[0135] Step 3.3, adjust the light field generation algorithm: Based on the calculated light field error, the next step is to adjust the light field generation algorithm so that the light field distribution tends to the target light field distribution and achieve parallax matching. The light field adjustment formula is as follows:

[0136]

[0137] Among them, L adjusted (x, y) is the light field distribution after adjustment, L current (x, y) is the current light field distribution, L target (x, y) is the target light field distribution, α is the adjustment coefficient;

[0138] Step S4: driving the variable focus liquid crystal lens array, controlling the refractive index of the liquid crystal lens on the basis of adjusting the light field distribution of the display screen, and realizing dynamic focal length adjustment;

[0139] Step 4.1, calculate the refractive index of the liquid crystal lens: After the light field control is completed, calculate the refractive index of the liquid crystal lens at different voltages. The refractive index n(V) of the liquid crystal lens is controlled by the electric field. The calculation formula is as follows:

[0140]

[0141] Where n(V) is the refractive index of the liquid crystal lens under the action of voltage V, n o is the normal refractive index of the liquid crystal material, n e is the anomalous refractive index of the liquid crystal material, V is the voltage applied to the liquid crystal lens, V t is the characteristic voltage that the liquid crystal molecules respond to;

[0142] Step 4.2, calculate the focal length of the liquid crystal lens: After calculating the refractive index of the liquid crystal lens, calculate the focal length f(V) based on the refractive index to determine the imaging characteristics of the liquid crystal lens at different voltages. The formula for calculating the focal length of the liquid crystal lens is as follows:

[0143] Where f(V) is the focal length of the liquid crystal lens under the action of voltage V, r is the radius of curvature of the liquid crystal lens, and n(V) is the refractive index of the liquid crystal lens calculated in step 4.1;

[0144] Step 4.3: Drive the liquid crystal lens array. After calculating the refractive index and focal length of the liquid crystal lens, the next step is to control the voltage of the liquid crystal lens array so that the lens focal length can match the optimal parallax distribution in real time. The driving control equation of the liquid crystal lens is as follows:

[0145]

[0146] Among them, V control (x, y) is the control voltage of the lens array at position (x, y), V min 、V max is the minimum and maximum operating voltage of the lens array, f target (x,y) is the target focal length, f min 、f max are the minimum and maximum focal lengths of the liquid crystal lens;

[0147] Step S5: adaptively adjust the display content. After the liquid crystal lens array completes dynamic focal length adjustment, the display content is optimized and rendered to match the current optimal parallax distribution. The process then returns to step S1 to continue user viewpoint detection and parameter acquisition, forming a real-time closed-loop control.

[0148] Step 5.1, optimize 3D image rendering parameters: After the focal length of the liquid crystal lens is adjusted, optimize the rendering parameters of the 3D image so that the 3D image can match the optimal parallax distribution δ optimal (x,y) and the focal length of the liquid crystal lens f(V), the calculation method is as follows:

[0149] Depth of Field Adjustment

[0150]

[0151] in, is the optimized depth of field value, D depth (x, y) is the depth map of the original display content, δ optimal (x,y) is the optimal disparity distribution, δ theory (x, y) is the theoretical parallax distribution, λ is the depth compensation coefficient;

[0152] Hierarchical Hybrid Optimization blend (x,y):

[0153] L blend (x,y)=w f ·L foreground (x,y)+w b ·L background (x,y),

[0154] Among them, L blend (x, y) is the brightness value of the mixed image, L foreground (x,y) and L background (x,y) represents the brightness distribution of the foreground and background, w f 、w b is the weight coefficient of foreground and background;

[0155] Focal length compensation f compensated (x,y):

[0156]

[0157] Among them, f compensated (x,y) is the focal length after compensation, f(V) is the focal length of the liquid crystal lens, D obs (x,y) is the observation distance, μ is the focal length compensation factor;

[0158] Step 5.2, adjust the inter-frame interpolation algorithm: After optimizing the 3D image rendering parameters, adjust the inter-frame interpolation algorithm to improve the smoothness of dynamic 3D display and reduce the visual delay caused by parallax adjustment and light field changes;

[0159] Inter-frame interpolation calculation:

[0160] Inew (x, y, t) = (1-β)·I prev (x,y,t)+β·I next (x,y,t),

[0161] Among them, I new (x, y, t) is the interpolated frame image, I prev (x,y,t) and I next (x, y, t) represents the image data of the previous frame and the next frame, and β is the interpolation weight;

[0162] Step 5.3: Use GPU parallel computing to accelerate 3D rendering: After completing image rendering parameter optimization and inter-frame interpolation adjustment, in order to improve the real-time performance of the system, use GPU parallel computing to accelerate the 3D rendering process;

[0163] Parallel computing framework:

[0164] Among them, T render is the rendering time after GPU parallel computing, T CPU N is the rendering time of CPU serial calculation, GPU Calculate the number of cores for the GPU.

[0165] Step S1: Benefits of user viewpoint detection and parameter acquisition: The user's eye coordinates and line of sight direction are obtained in real time through the eye tracking device, which improves the accuracy of parallax calculation; the 3D parallax is dynamically adjusted according to the user's observation distance and pupil distance to ensure that the image conforms to the visual characteristics of the human eye; different users have different eye distances, viewing angles, and screen distances, which achieves personalized adaptation and improves 3D viewing comfort.

[0166] Step S2: Benefits of dynamic optimal disparity matching calculation: The optimal disparity distribution is calculated based on the user's viewpoint information and the 3D content depth information, making the picture immersive. The Euler-Lagrange variational optimization method is used to ensure smooth transition of disparity adjustment to avoid discomfort caused by jumps. GPU parallel computing accelerates the optimal disparity solution to ensure that the 3D picture can adapt to the user's head movement or viewing angle changes in real time.

[0167] Step S3: Benefits of light field control: By calculating the target light field distribution, the light field is matched with the optimal parallax, thereby improving the stereoscopic effect; by calculating the light field error, the light field generation algorithm is optimized, so that the light field output by the display is close to the theoretical optimal light field; the optimized light field reduces common 3D visual artifacts, making the 3D image clear.

[0168] Step S4: Benefits of driving a variable-focus liquid crystal lens array: The refractive index of the liquid crystal lens is adjustable, and the focal length is dynamically adjusted by environmental changes, thereby enhancing the depth perception of the 3D image. Compared with MEMS lenses, liquid crystal lenses have a fast response speed and can complete focal length adjustment within milliseconds. Traditional lenses have a fixed focal length and cannot adapt to different parallax requirements, while liquid crystal lenses can be adjusted by changes in display content and user position.

[0169] Step S5: Benefits of adaptive adjustment of display content: Through depth of field adjustment, focal length compensation and other technologies, the 3D picture is made natural and distortion is reduced; the inter-frame interpolation algorithm is optimized to reduce the screen tearing during parallax adjustment and improve the smoothness of dynamic display; parallel computing shortens rendering time, so that the 3D picture adjustment can respond to changes in user perspective in real time.

[0170] In summary, the control method of the naked-eye 3D display device of the present invention realizes the full-process closed-loop control of user viewpoint detection, dynamic parallax matching, light field regulation, liquid crystal lens focal length adjustment and display content optimization. Compared with the traditional naked-eye 3D solution, the technical solution of the present invention can dynamically adapt to different users, improve the comfort of 3D viewing, and reduce visual fatigue. At the same time, the optimization process is accelerated by GPU parallel computing, which significantly reduces the computing delay and makes the 3D picture smooth. Overall, the present invention improves the adaptability, clarity and immersion of naked-eye 3D display technology, and provides efficient and real-time solutions for future 3D interactive displays and AR / VR devices.

[0171] A terminal device includes a memory, a processor and a display device, wherein the processor is used to execute a control method for a naked-eye 3D display device.

[0172] A storage medium stores computer-readable instructions. When the instructions are executed by a terminal device, the terminal device executes a control method for a naked-eye 3D display device.

[0173] It supports efficient storage of 3D rendering data, reduces data reading latency, and improves the real-time performance of 3D display; stores intermediate calculation results, reduces the GPU computing burden, and improves the efficiency of parallax optimization calculations; ensures the parallel execution of viewpoint tracking, light field calculation, and liquid crystal lens control tasks, and improves system response speed.

[0174] It uses a high-performance CPU and GPU, supports parallel computing, and accelerates dynamic parallax matching and light field control. It supports AI-based eye tracking and 3D parallax optimization to improve user adaptability. It optimizes computing resource allocation, reduces power consumption of 3D rendering and parallax calculation, and improves device battery life.

[0175] Combined with a variable-focus liquid crystal lens array, dynamic focal length adjustment is achieved to enhance the 3D stereoscopic effect; through dynamic optimal parallax matching, the 3D image is made to conform to the visual habits of the human eye, reducing viewing fatigue; and the problems of image tearing and ghosting are reduced, improving the 3D dynamic display effect.

[0176] In summary, the terminal device of the present invention combines high-performance memory, a processor, and an intelligent display device to achieve full-process optimization for naked-eye 3D display. Compared to traditional devices, the present invention significantly improves computing efficiency, visual comfort, and system adaptability. GPU parallel computing accelerates 3D rendering, and the liquid crystal lens array dynamically adjusts the light field, resulting in clear and smooth 3D images. Furthermore, the device optimizes power consumption management to ensure stable 3D viewing even during extended periods.

[0177] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for controlling a naked-eye 3D display device, characterized in that: include: Step S1: user viewpoint detection and parameter acquisition, using sensors to collect the user's eye position, viewing angle, and viewing distance, and calculating theoretical parallax values based on the acquired parameters; Step S2: Dynamic optimal disparity matching calculation, based on the user's eye position, viewing angle and viewing distance, combined with the depth information of the displayed content, calculates the optimal disparity distribution in the current environment; Step S3: light field control. After calculating the optimal parallax distribution, the light field distribution of the display screen is adjusted according to the distribution. Step S4: driving the variable focus liquid crystal lens array, controlling the refractive index of the liquid crystal lens on the basis of adjusting the light field distribution of the display screen, and realizing dynamic focal length adjustment; Step S5, adaptively adjust the display content. After the liquid crystal lens array completes the dynamic focal length adjustment, the display content is optimized and rendered to match the current optimal parallax distribution. Then, the system returns to step S1 to continue user viewpoint detection and parameter acquisition, forming a real-time closed-loop control.

2. The control method of a naked-eye 3D display device according to claim 1, characterized in that: In step S1, user viewpoint detection and parameter acquisition further include: Step 1.1, user eye tracking: Use the eye tracking device to obtain the user's eye position and gaze direction in real time, and set the eye tracking device to output the current eye position vector and gaze direction vector; Eye position: P eye =(x eye ,y eye ,z eye ), Among them, P eye is the position coordinate of the user's eyes, x eye 、y eye 、z eye Represents the coordinates of the user's eyes in three-dimensional space; Sight direction: D view =(d x ,d y ,d z ), Among them, D view is the direction vector of the user's line of sight, d x d y d z Represents the components of the sight direction on the x, y, and z coordinate axes; Step 1.2, calculate the viewing distance and pupil distance: Calculate the viewing distance D by using the eye position and line of sight obtained by the eye tracking device and combining it with the known geometric parameters of the display. obs and pupil distance P interpupillary ; Among them, (x screen ,y screen ,z screen ) is the spatial coordinate of a point on the display screen, D obs is the observation distance; P interpupillary =|x left -x right |, Among them, P interpupillary is the pupil distance, x left 、x right are the position coordinates of the left eye and the right eye; Step 1.3, generate theoretical disparity distribution: calculate theoretical disparity δ using the Horopter curve theory The Horopter curve describes the trajectory of the point with the smallest parallax of the human eye. The formula is as follows: Among them, δ theory is the theoretical parallax, b z is the depth of the object; Step 1.4, obtain the display content depth map: used to obtain the depth information of the current 3D display content, set the depth map of the display content to D depth (x, y) represents the depth value of each pixel in the display content; Depth Map D depth (x,y):D depth (x, y) is the pixel depth at position (x, y) in the display coordinate system, which is extracted from the 3D model using a computer vision algorithm and represents the relative distance of objects in the displayed content.

3. The control method of a naked-eye 3D display device according to claim 1, characterized in that: In step S2, the dynamic optimal disparity matching calculation further includes: Step 2.1, set the objective function and solve the optimal disparity distribution: To optimize the disparity matching effect, set the objective function. The calculated optimal disparity distribution can meet the requirements of minimizing the visual fusion error, image content constraint error, temporal continuity error, and light field optimization error. The objective function is defined as follows: E(δ)=w1E fusion (d)+w2E content (d)+w3E temporal (d)+w4E lightfield (d), Among them, E(δ) is the total optimization objective function of the optimal disparity distribution, E fusion (δ) is the visual fusion error, E content (δ) is the image content constraint error, E temporal (δ) is the time continuity error, E lightfield (δ) is the light field optimization error, w1, w2, w3, w4 are the weight coefficients of each error term; Step 2.2, establish the parallax optimization equation: After the objective function is set, build a parallax optimization solution model based on the Euler-Lagrange equation. The expression of the Euler-Lagrange equation is as follows: Among them, δ is the currently calculated disparity distribution, x and y are the spatial coordinates on the display screen, t is the time, and δ x , δ y is the partial derivative of the disparity distribution in the spatial direction, δ t is the partial derivative of the disparity distribution in time, and d is a constant term.

4. The control method of a naked-eye 3D display device according to claim 1, characterized in that: In step S2, the dynamic optimal disparity matching calculation further includes: Step 2.3: Solve using numerical optimization methods: Since the analytical solution of the Euler-Lagrange equation is relatively complex, a numerical optimization method is used to calculate to obtain the optimal disparity distribution. Specific methods include the gradient descent method and the conjugate gradient method. The calculation formula of the gradient descent method is as follows: Among them, δ (n) is the disparity distribution of the nth iteration, η is the learning rate, is the objective function gradient under the current disparity distribution, δ (n+1) is the disparity distribution of the nth iteration; If the conjugate gradient method is used, the update formula is as follows: d (n+1) =d (n) +αp (n) , Among them, p (n) is the conjugate direction of the nth iteration, α is the step size factor, β is the conjugate coefficient, is the objective function gradient under the parallax distribution at the next moment, p (n+1) is the conjugate direction of the n+1th iteration; After completing the numerical optimization calculation, the final optimal disparity distribution is obtained, and the optimal disparity distribution will be used for light field control in step S3.

5. The control method of a naked-eye 3D display device according to claim 1, characterized in that: In step S3, light field control further includes: Step 3.1, calculate the target light field distribution: After obtaining the optimal disparity distribution, the next step is to calculate the target light field distribution based on the disparity optimization results. The target light field distribution L target (x,y) is the optimal disparity distribution δ calculated based on optimal (x,y) and display content depth map D depth The relationship between (x, y) is obtained; Target light field distribution L target (x,y) is given by: Among them, L target (x, y) is the target light field distribution, D depth (x, y) is the depth value of the display content at position (x, y), δ optimal (x,y) is the optimal disparity distribution, D obs (x, y) is the viewing distance between the user and the display screen; Step 3.2, calculate the light field error: After calculating the target light field distribution, evaluate the deviation between the current display light field and the target light field, and define the light field error E f (D) is used to measure the difference between the current light field and the target light field. The light field error formula is as follows: E f (D)=∑ x,y (L current (x,y)-L target (x,y)) 2 , Among them, E f (D) is the light field error, L current is the light field distribution of the current display, L target (x,y) is the target light field distribution; Step 3.3, adjust the light field generation algorithm: Based on the calculated light field error, the next step is to adjust the light field generation algorithm so that the light field distribution tends to the target light field distribution and achieve parallax matching. The light field adjustment formula is as follows: Among them, L adjusted (x, y) is the light field distribution after adjustment, L current (x, y) is the current light field distribution, L target (x, y) is the target light field distribution, and α is the adjustment coefficient.

6. The control method of a naked-eye 3D display device according to claim 1, characterized in that: In step S4, driving the zoom liquid crystal lens array further includes: Step 4.1, calculate the refractive index of the liquid crystal lens: After the light field control is completed, calculate the refractive index of the liquid crystal lens at different voltages. The refractive index n(V) of the liquid crystal lens is controlled by the electric field. The calculation formula is as follows: Where n(V) is the refractive index of the liquid crystal lens under the action of voltage V, n o is the normal refractive index of the liquid crystal material, n e is the anomalous refractive index of the liquid crystal material, V is the voltage applied to the liquid crystal lens, V t is the characteristic voltage of the liquid crystal molecules; Step 4.2, calculate the focal length of the liquid crystal lens: After calculating the refractive index of the liquid crystal lens, calculate the focal length f(V) based on the refractive index to determine the imaging characteristics of the liquid crystal lens at different voltages. The formula for calculating the focal length of the liquid crystal lens is as follows: Where f(V) is the focal length of the liquid crystal lens under the action of voltage V, r is the radius of curvature of the liquid crystal lens, and n(V) is the refractive index of the liquid crystal lens calculated in step 4.1; Step 4.3: Drive the liquid crystal lens array. After calculating the refractive index and focal length of the liquid crystal lens, the next step is to control the voltage of the liquid crystal lens array so that the lens focal length can match the optimal parallax distribution in real time. The driving control equation of the liquid crystal lens is as follows: Among them, V control (x, y) is the control voltage of the lens array at position (x, y), V min 、V max is the minimum and maximum operating voltage of the lens array, f target (x,y) is the target focal length, f min 、f max are the minimum and maximum focal lengths of the liquid crystal lens.

7. The control method of a naked-eye 3D display device according to claim 1, characterized in that: In step S5, the adaptive adjustment of the display content further includes: Step 5.1, optimize 3D image rendering parameters: After the focal length of the liquid crystal lens is adjusted, optimize the rendering parameters of the 3D image so that the 3D image can match the optimal parallax distribution δ optimal (x,y) and the focal length of the liquid crystal lens f(V), the calculation method is as follows: Depth of Field Adjustment in, is the optimized depth of field value, D depth (x, y) is the depth map of the original display content, δ optimal (x,y) is the optimal disparity distribution, δ theory (x, y) is the theoretical parallax distribution, λ is the depth compensation coefficient; Hierarchical Hybrid Optimization blend (x,y): L blend (x,y)=w f ·L foreground (x,y)+w b ·L background (x,y), Among them, L blend (x, y) is the brightness value of the mixed image, L foreground (x,y) and L background (x,y) represents the brightness distribution of the foreground and background, w f 、w b is the weight coefficient of foreground and background; Focal length compensation f compensated (x,y): Among them, f compensated (x,y) is the focal length after compensation, f(V) is the focal length of the liquid crystal lens, D obs (x,y) is the observation distance, and μ is the focal length compensation factor.

8. The control method of a naked-eye 3D display device according to claim 1, characterized in that: In step S5, the adaptive adjustment of the display content further includes: Step 5.2, adjust the inter-frame interpolation algorithm: After optimizing the 3D image rendering parameters, adjust the inter-frame interpolation algorithm to improve the smoothness of dynamic 3D display and reduce the visual delay caused by parallax adjustment and light field changes; Inter-frame interpolation calculation: I new (x,y,t)=(1-β)·I prev (x,y,t)+β·I next (x,y,t), Among them, I new (x, y, t) is the interpolated frame image, I prev (x,y,t) and I next (x, y, t) represents the image data of the previous frame and the next frame, and β is the interpolation weight; Step 5.3: Use GPU parallel computing to accelerate 3D rendering: After completing image rendering parameter optimization and inter-frame interpolation adjustment, in order to improve the real-time performance of the system, use GPU parallel computing to accelerate the 3D rendering process; Parallel computing framework: Among them, T render is the rendering time after GPU parallel computing, T CPU N is the rendering time of CPU serial calculation, GPU Calculate the number of cores for the GPU.

9. A terminal device, characterized in that: The device comprises a memory, a processor and a display device, wherein the processor is used to execute the control method of the naked-eye 3D display device according to any one of claims 1 to 8.

10. A storage medium, characterized in that: The storage medium stores computer-readable instructions, and when the instructions are executed by a terminal device, the terminal device executes the control method of a naked-eye 3D display device according to any one of claims 1 to 8.

Images