Phase Modulation Method, System, Device and Medium of a Liquid Crystal Lens Array

Abstract

The present invention discloses a phase modulation method, system, device and medium for a liquid crystal lens array, which relates to the technical field of liquid crystal device modulation. First, obtain the optical illuminance distribution of a three-dimensional display system under each electric field distribution, construct a data set, and use the data set to train an initial prediction model to obtain a trained prediction model. Then, using the target optical illuminance distribution as the input, determine the target electric field distribution by using the trained prediction model, and load the target electric field distribution on the liquid crystal lens array to perform phase modulation on the liquid crystal lens array. Thus, through the neural network, it is only necessary to make one prediction to determine the electric field distribution of the liquid crystal lens array corresponding to the target optical illuminance distribution, so as to achieve phase modulation through electric field modulation. The modulation is simple, and the optical display effect after modulation can make the optical crosstalk reach a lower level.

Description

Technical Field

[0001] The present invention relates to the technical field of liquid crystal device modulation, and particularly to a phase modulation method, system, device and medium for a liquid crystal lens array based on a neural network to reduce optical crosstalk in 3D display. Background Art

[0002] In recent years, 3D display technology based on liquid crystal lens arrays has gradually become the mainstream technology of 3D display technology due to the variable focal length attribute of liquid crystal lenses without mechanical movement. Limited by the liquid crystal device processing technology, as well as the requirements for the driving voltage and response speed of liquid crystal devices, the gradient liquid crystal lens cannot meet the requirements of the zooming ability. Therefore, liquid crystal Fresnel lenses with multiple electrodes are produced. However, for a liquid crystal lens array based on a liquid crystal Fresnel lens, since it involves multi-electrode modulation and the electrode scale is too small, edge field effects will occur, and the precise modulation of the phase distribution is often very difficult. Moreover, as the size of the liquid crystal lens array increases or the electrode scale of the electrode device decreases, the modulation difficulty will also increase accordingly. At the same time, after multiple electric field modulations in the liquid crystal lens array, the obtained phase distribution result cannot achieve a good optical display effect.

[0003] Based on this, there is an urgent need for a phase modulation technology with simple modulation and an optical display effect after modulation that can make the optical crosstalk lower. Summary of the Invention

[0004] The purpose of the present invention is to provide a phase modulation method, system, device and medium for a liquid crystal lens array, which determines the electric field distribution of the liquid crystal lens array corresponding to the target optical illuminance distribution through a neural network, so as to realize phase modulation through electric field modulation, with simple modulation and an optical display effect after modulation that can make the optical crosstalk lower.

[0005] To achieve the above purpose, the present invention provides the following solutions:

[0006] A phase modulation method for a liquid crystal lens array, the phase modulation method comprising:

[0007] Obtain the optical illuminance distribution of a 3D display system under each electric field distribution, and construct a data set; the electric field distribution is loaded on the liquid crystal lens array of the 3D display system; the data set includes a plurality of the electric field distributions and the optical illuminance distribution under each electric field distribution;

[0008] Train an initial prediction model using the data set to obtain a trained prediction model;

[0009] Taking the target optical illuminance distribution as the input, using the trained prediction model to determine the target electric field distribution, and loading the target electric field distribution on the liquid crystal lens array to perform phase modulation on the liquid crystal lens array.

[0010] In some embodiments, the liquid crystal lens array includes an upper substrate, a common electrode, an alignment layer, a liquid crystal layer, an alignment layer, an ITO electrode, and a lower substrate that are stacked in sequence from top to bottom.

[0011] In some embodiments, the trained prediction model uses a hybrid neural network model; the hybrid neural network model includes a convolutional neural network and a recurrent neural network.

[0012] In some embodiments, the hybrid neural network model includes a convolutional neural network and a recurrent neural network connected in parallel, a fusion layer connected to both the output of the convolutional neural network and the output of the recurrent neural network, a fully connected layer connected to the fusion layer, and an output layer connected to the fully connected layer.

[0013] In some embodiments, the input of the convolutional neural network is the optical illuminance distribution; the input of the recurrent neural network is the feature of the optical illuminance distribution; the feature is the illuminance peak and optical crosstalk.

[0014] In some embodiments, the convolutional neural network includes a convolutional layer, a pooling layer, and a fully connected layer; the recurrent neural network includes a long short-term memory network layer and a fully connected layer, or the recurrent neural network includes a gated recurrent unit network and a fully connected layer.

[0015] A phase modulation system for a liquid crystal lens array, the phase modulation system comprising:

[0016] A dataset construction module for obtaining the optical illuminance distribution of a three-dimensional display system under each electric field distribution and constructing a dataset; the electric field distribution is loaded on the liquid crystal lens array of the three-dimensional display system; the dataset includes a plurality of the electric field distributions and the optical illuminance distribution under each electric field distribution;

[0017] A training module for training an initial prediction model using the dataset to obtain a trained prediction model;

[0018] A modulation module for taking the target optical illuminance distribution as the input, using the trained prediction model to determine the target electric field distribution, and loading the target electric field distribution on the liquid crystal lens array to perform phase modulation on the liquid crystal lens array.

[0019] In some embodiments, the trained prediction model uses a hybrid neural network model; the hybrid neural network model includes a convolutional neural network and a recurrent neural network.

[0020] A phase modulation device for a liquid crystal lens array, comprising:

[0021] A processor; and

[0022] A memory storing computer-readable program instructions,

[0023] wherein, when the computer-readable program instructions are run by the processor, the above-mentioned phase modulation method is executed.

[0024] A computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned phase modulation method are implemented.

[0025] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0026] The present invention is used to provide a phase modulation method, system, device and medium for a liquid crystal lens array. First, the optical illuminance distribution of a three-dimensional display system under each electric field distribution is obtained, a data set is constructed, and the initial prediction model is trained using the data set to obtain a trained prediction model. Then, with the target optical illuminance distribution as the input, the trained prediction model is used to determine the target electric field distribution, and the target electric field distribution is loaded on the liquid crystal lens array to perform phase modulation on the liquid crystal lens array. Thus, through the neural network, only one prediction is required to determine the electric field distribution of the liquid crystal lens array corresponding to the target optical illuminance distribution, so as to achieve phase modulation through electric field modulation. The modulation is simple, and the optical display effect after modulation can make the optical crosstalk reach a lower level. Description of the Drawings

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0028] Figure 1 It is the method flow chart of the phase modulation method provided by Embodiment 1 of the present invention;

[0029] Figure 2 It is the principle block diagram of the phase modulation method provided by Embodiment 1 of the present invention;

[0030] Figure 3 It is the optical path schematic diagram of the lens-type three-dimensional display principle provided by Embodiment 1 of the present invention;

[0031] Figure 4Schematic diagram of the optical illuminance distribution provided by Embodiment 1 of the present invention;

[0032] Figure 5 Schematic diagram of the structure of the liquid crystal lens array provided by Embodiment 1 of the present invention;

[0033] Figure 6 Schematic diagram of the structure of the hybrid neural network model provided by Embodiment 1 of the present invention;

[0034] Figure 7 System block diagram of the phase modulation system provided by Embodiment 2 of the present invention. Detailed implementation manners

[0035] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0036] The object of the present invention is to provide a phase modulation method, system, device and medium for a liquid crystal lens array, which determines the electric field distribution of the liquid crystal lens array corresponding to the target optical illuminance distribution through a neural network, so as to achieve phase modulation through electric field modulation. The modulation is simple, and the optical display effect after modulation can make the optical crosstalk reach a lower level.

[0037] To make the above objects, features and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0038] Embodiment 1:

[0039] By applying voltages to multiple drive electrodes of a liquid crystal lens array, the liquid crystal lens array is subjected to electric field modulation to form a corresponding electric field distribution in the liquid crystal lens array. The non-uniform electric field induces the liquid crystal molecules in different regions of the liquid crystal lens array to deflect at different angles, and the director of the liquid crystal molecules changes, resulting in a phase delay distribution in the liquid crystal lens array. That is, the deflection of the liquid crystal molecules in the liquid crystal lens array driven by this electric field forms a phase difference, enabling the liquid crystal lens array to have the phase distribution of a lens. Under this phase distribution, a three-dimensional display system including this liquid crystal lens array will generate an optical illuminance distribution, which refers to the horizontal luminance distribution at the optimal three-dimensional display viewing distance in three-dimensional display. The traditional phase modulation method is to perform electric field modulation manually. After each modulation, an optical illuminance distribution is obtained, and then the modulation effect is judged, and the next electric field modulation is carried out according to the modulation effect. However, the optical illuminance distribution obtained after multiple modulations is often not the most perfect. That is to say, more refined electric field modulation is required to achieve a better optical display effect, and higher modulation accuracy means more complex modulation steps. Therefore, the traditional phase modulation method has problems of low modulation efficiency and poor modulation effect.

[0040] To solve the above problems, this embodiment is used to provide a phase modulation method for a liquid crystal lens array, as Figure 1 and Figure 2 shown. The phase modulation method includes:

[0041] S1: Obtain the optical illuminance distribution of the three-dimensional display system under each electric field distribution and construct a data set; the electric field distribution is loaded on the liquid crystal lens array of the three-dimensional display system; the data set includes multiple electric field distributions and the optical illuminance distribution under each electric field distribution;

[0042] As Figure 3 shown, the three-dimensional display system of this embodiment includes a two-dimensional display screen, a liquid crystal lens array, and viewpoints 1 and 2 as the human eyes. The distance between viewpoints 1 and 2 is the interpupillary distance of the human eyes. Of course, the three-dimensional display system of this embodiment may also include multiple other viewpoints, and all viewpoints are in the same plane. The liquid crystal lens array is located directly above the two-dimensional display screen, and the position of the two-dimensional display screen is on the focal plane of the liquid crystal lens array. The three-dimensional display process of this three-dimensional display system includes: the light emitted by the pixels in the two-dimensional display screen passes through the refraction of the liquid crystal lens array and converges to different viewpoints. Electric field modulation of the liquid crystal lens array can make the liquid crystal lens array play a focusing role, and the observer realizes three-dimensional perception through binocular parallax, thereby forming a three-dimensional display effect.

[0043] In this embodiment, the electrode voltage of the liquid crystal lens array is first modulated, that is, electric field modulation is performed to form an electric field distribution. The electric field modulation process includes: applying a voltage to the driving electrodes of the liquid crystal lens array to form an electric field distribution in the liquid crystal lens array. Since the liquid crystal lens array applied in this embodiment is a multi-electrode structure, different voltage application schemes can be obtained by applying different voltages to different driving electrodes. Each voltage application scheme includes the voltage applied to each driving electrode. Different electric field distributions will be generated in the liquid crystal lens array under different voltage application schemes. Then, the modulated liquid crystal lens array is used in a 3D display system to test its display effect. The display effect of the liquid crystal lens array in the 3D display system is measured by the optical illuminance distribution. The optical illuminance distribution includes the brightness distribution of the pixels received at different viewpoints, including the brightness information of the pixels in the 2D display screen received at different viewpoints. As Figure 4 shown, when there are 9 viewpoints, the optical illuminance distribution includes the brightness information of the pixels in the 2D display screen received at 9 viewpoints, Figure 4 which are the brightness distributions corresponding to 9 viewpoints (viewpoint 1 - viewpoint 9) from left to right. Multiple groups of electric field distributions and the optical illuminance distributions under each electric field distribution are obtained by modulating the electric field of the liquid crystal lens array multiple times. That is, in this embodiment, different electric field distributions are loaded onto the liquid crystal lens array multiple times to obtain the optical illuminance distribution of the 3D display system under each electric field distribution. By collecting the optical illuminance distributions under different electric field distributions, a data set for training the initial prediction model is constructed. This data set contains multiple different electric field distributions of the liquid crystal lens array and the optical illuminance distribution of the corresponding 3D display system under each electric field distribution.

[0044] Preferably, as Figure 5 shown, the liquid crystal lens array of this embodiment includes an upper substrate, a common electrode, an alignment layer, a liquid crystal layer, an alignment layer, an ITO electrode (i.e., a driving electrode), and a lower substrate that are stacked in sequence from top to bottom. Among them, the materials of the upper substrate and the lower substrate are generally glass, and can also be transparent materials such as plastics, such as PET. That is, the upper substrate of the liquid crystal lens array in this embodiment can use an upper glass substrate, and the lower substrate can use a lower glass substrate. The material of the common electrode is generally a transparent conductive metal oxide, such as ITO (indium tin oxide), etc., and can also be a transparent conductive organic polymer material, such as PEDOT (polyethylene dioxythiophene), etc., or an inorganic material, such as a metal grid, silver nanowire, graphene, and carbon nanotube, etc. The alignment method of the alignment layer can be rubbing alignment or ultraviolet exposure. That is, the alignment layer of the liquid crystal lens array in this embodiment can use a rubbing alignment layer. The material of the liquid crystal layer can be any liquid crystal material.

[0045] S2: Use the data set to train the initial prediction model to obtain a trained prediction model;

[0046] The initial prediction model in this embodiment adopts a hybrid neural network model, which includes a convolutional neural network and a recurrent neural network. That is, in this embodiment, a hybrid neural network model with a parallel convolutional neural network and recurrent neural network is constructed as the initial prediction model. The data set is used as the input of the initial prediction model, and the initial prediction model is trained to enable the initial prediction model to learn the electric field distribution of the liquid crystal lens array under different display effects (optical illumination distribution). After training is completed, a trained prediction model can be obtained. Therefore, the trained prediction model in this embodiment adopts a hybrid neural network model, which includes a convolutional neural network and a recurrent neural network.

[0047] As Figure 6 shown, the hybrid neural network model constructed in this embodiment for modulating the liquid crystal lens array includes a convolutional neural network and a recurrent neural network connected in parallel, a fusion layer connected to both the output of the convolutional neural network and the output of the recurrent neural network, a fully connected layer connected to the fusion layer, and an output layer connected to the fully connected layer. That is, the hybrid neural network model in this embodiment performs parallel output by the convolutional neural network and the recurrent neural network. Specifically, the outputs of the two neural networks are fused and output through a fusion layer (i.e., the Concatenate layer). The outputs of the two neural networks are both tensors. By inputting the tensors of the outputs of the two neural networks into the fusion layer, the fusion layer outputs an output tensor formed by connecting all the input tensors. The output tensor then enters the subsequent fully connected layer and output layer to obtain the electric field distribution of the liquid crystal lens array.

[0048] Among them, the input of the convolutional neural network is the optical illumination distribution. The convolutional neural network includes a convolutional layer, a pooling layer, and a fully connected layer. The convolutional layer can include multiple layers, and the pooling layer can be a max pooling layer. Specifically, as an example, the convolutional neural network can include six consecutive convolutional pooling blocks and one fully connected layer. Each convolutional pooling block includes a convolutional layer and a pooling layer connected in sequence.

[0049] The input of the recurrent neural network is the feature of the optical illumination distribution, aiming to capture the correlation between different groups of data in the data set. In this embodiment, any viewpoint is selected, and the feature of this viewpoint is used as the feature of the optical illumination distribution. The feature of the viewpoint can be the illumination peak value and optical crosstalk. The illumination peak value is the maximum brightness value in the brightness distribution received by this viewpoint, and the optical crosstalk is the ratio of the stray light in the brightness distribution received by this viewpoint to the sum of the illumination peak value and the stray light. Among them, crosstalk is the optical crosstalk, P stray is the stray light, P normalis the peak illuminance. The stray light is determined as follows: determine the corresponding value at the position corresponding to the peak illuminance of the brightness distribution of each other viewing point at this viewing point, and add the corresponding values of each other viewing point to obtain the stray light. The recurrent neural network may include a long short-term memory network layer and a fully connected layer. Of course, the long short-term memory network layer may also be other recurrent neural networks such as a gated recurrent unit network, etc. Therefore, the recurrent neural network in this embodiment may also include a gated recurrent unit network and a fully connected layer. More specifically, as an example, the recurrent neural network in this embodiment includes three sequentially connected long short-term memory network layers and one fully connected layer, or the recurrent neural network in this embodiment includes three sequentially connected gated recurrent unit network layers and one fully connected layer.

[0050] S3: Using the target optical illuminance distribution as the input, determine the target electric field distribution by using the trained prediction model, and load the target electric field distribution on the liquid crystal lens array to perform phase modulation on the liquid crystal lens array.

[0051] The target optical illuminance distribution in this embodiment is determined according to the required ideal display effect. The ideal display effect means that there is no optical crosstalk, that is, the display effect without stray light. This target optical illuminance distribution can be generated by simulation and is the optical illuminance distribution without optical crosstalk. In this embodiment, the optical illuminance distribution under an ideal gradient lens can also be used as the target optical illuminance distribution.

[0052] Using the target optical illuminance distribution as the input, through one prediction by the trained prediction model, the target electric field distribution of the liquid crystal lens array that can achieve the ideal display effect can be obtained. After loading the predicted target electric field distribution on the liquid crystal lens array, the liquid crystal molecules are induced to deflect to form a new phase distribution, that is, phase modulation is performed, and under the new phase distribution, a better optical display effect can be obtained, such as a higher peak illuminance and lower optical crosstalk.

[0053] A phase modulation method for a liquid crystal lens array applicable to three-dimensional display proposed in this embodiment. The modulation method is to accumulate a data set through prior manual debugging, and then train a hybrid neural network model with a certain number of accumulated data sets to obtain a trained prediction model, so as to predict the new electric field distribution of the liquid crystal lens array under the target optical illuminance distribution. After loading the predicted electric field distribution onto the liquid crystal lens array, the liquid crystal molecules are induced to deflect to form a new phase distribution. The obtained phase distribution will have a better optical illuminance distribution when applied to a three-dimensional display system, that is, a better optical display effect (three-dimensional display effect) can be achieved under the new phase distribution, such as a higher illuminance peak and lower optical crosstalk. The electric field modulation of the liquid crystal lens is realized through a neural network, and further the phase modulation of the liquid crystal lens is realized. Compared with the existing phase modulation technology, this modulation method has higher efficiency, and the optical display effect achieved after modulation can make the optical crosstalk reach a lower level.

[0054] Embodiment 2:

[0055] This embodiment is used to provide a phase modulation system for a liquid crystal lens array, as Figure 7 shown. The phase modulation system includes:

[0056] A data set construction module M1, which is used to obtain the optical illuminance distribution of the three-dimensional display system under each electric field distribution and construct a data set; the electric field distribution is loaded on the liquid crystal lens array of the three-dimensional display system; the data set includes a plurality of the electric field distributions and the optical illuminance distribution under each electric field distribution;

[0057] A training module M2, which is used to train an initial prediction model with the data set to obtain a trained prediction model;

[0058] A modulation module M3, which is used to take the target optical illuminance distribution as an input, use the trained prediction model to determine the target electric field distribution, and load the target electric field distribution on the liquid crystal lens array to perform phase modulation on the liquid crystal lens array.

[0059] The trained prediction model adopts a hybrid neural network model; the hybrid neural network model includes a convolutional neural network and a recurrent neural network.

[0060] This embodiment uses the above-mentioned proposed hybrid neural network model to train the hybrid neural network model through the constructed data set to obtain a trained prediction model, and predicts the target electric field distribution of the liquid crystal lens array corresponding to the target optical illuminance distribution through the trained prediction model. Then, the predicted target electric field distribution is loaded onto the liquid crystal lens array, and the obtained phase distribution will have a better optical illuminance distribution when applied to three-dimensional display.

[0061] Embodiment 3:

[0062] This embodiment is used to provide a phase modulation device for a liquid crystal lens array, including:

[0063] A processor; and

[0064] A memory, in which computer-readable program instructions are stored,

[0065] wherein, when the computer-readable program instructions are run by the processor, the phase modulation method described in Embodiment 1 is executed.

[0066] Embodiment 4:

[0067] This embodiment is used to provide a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the phase modulation method described in Embodiment 1 are implemented.

[0068] In each embodiment of this specification, the key points are the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0069] In this article, specific examples are used to elaborate on the principles and implementation manners of the present invention. The descriptions of the above embodiments are only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A phase modulation method for a liquid crystal lens array, characterized in that, the phase modulation method includes: obtaining the optical illuminance distribution of a three-dimensional display system under each applied electric field distribution, and constructing a data set; the electric field distribution is applied to the liquid crystal lens array of the three-dimensional display system; the data set includes a plurality of the electric field distributions and the optical illuminance distribution under each electric field distribution; using the data set to train an initial prediction model to obtain a trained prediction model; taking the target optical illuminance distribution as an input, using the trained prediction model to determine the target electric field distribution, and applying the target electric field distribution to the liquid crystal lens array to perform phase modulation on the liquid crystal lens array.

2. The phase modulation method according to claim 1, characterized in that, the liquid crystal lens array includes an upper substrate, a common electrode, an alignment layer, a liquid crystal layer, an alignment layer, an ITO electrode, and a lower substrate that are sequentially stacked from top to bottom.

3. The phase modulation method according to claim 1, characterized in that, the trained prediction model adopts a hybrid neural network model; the hybrid neural network model includes a convolutional neural network and a recurrent neural network.

4. The phase modulation method according to claim 3, characterized in that, the hybrid neural network model includes a convolutional neural network and a recurrent neural network connected in parallel, a fusion layer connected to both the output of the convolutional neural network and the output of the recurrent neural network, a fully connected layer connected to the fusion layer, and an output layer connected to the fully connected layer.

5. The phase modulation method according to claim 4, characterized in that, the input of the convolutional neural network is the optical illuminance distribution; the input of the recurrent neural network is the feature of the optical illuminance distribution; the feature is the illuminance peak and optical crosstalk.

6. The phase modulation method according to claim 4, characterized in that, the convolutional neural network includes a convolutional layer, a pooling layer, and a fully connected layer; the recurrent neural network includes a long short-term memory network layer and a fully connected layer, or the recurrent neural network includes a gated recurrent unit network and a fully connected layer.

7. A phase modulation system for a liquid crystal lens array, characterized in that, the phase modulation system includes: a data set construction module for obtaining the optical illuminance distribution of a three-dimensional display system under each applied electric field distribution, and constructing a data set; the electric field distribution is applied to the liquid crystal lens array of the three-dimensional display system; the data set includes a plurality of the electric field distributions and the optical illuminance distribution under each electric field distribution; a training module for using the data set to train an initial prediction model to obtain a trained prediction model; a modulation module for taking the target optical illuminance distribution as an input, using the trained prediction model to determine the target electric field distribution, and applying the target electric field distribution to the liquid crystal lens array to perform phase modulation on the liquid crystal lens array.

8. The phase modulation system according to claim 7, characterized in that, The trained prediction model adopts a hybrid neural network model; the hybrid neural network model includes a convolutional neural network and a recurrent neural network.

9. A phase modulation device for a liquid crystal lens array, characterized in that, it includes: a processor; and a memory, which stores computer-readable program instructions, wherein, when the computer-readable program instructions are run by the processor, the phase modulation method according to any one of claims 1-6 is executed.

10. A computer-readable storage medium, on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the steps of the phase modulation method according to any one of claims 1-6 are implemented.

Images