Decoupling control method for local dimming

Abstract

The decoupling control method for local dimming includes: analyzing the input image to obtain the target brightness of each dimming area, thereby establishing a target brightness matrix; inputting the target brightness matrix into a backlight decoupling neural network model to obtain the optimal dimming brightness matrix; and using the optimal dimming brightness to control the backlight module to emit light. The training dataset used in the modeling process of the backlight decoupling neural network model is related to a constrained least squares problem, the goal of which is to find the optimal dimming brightness matrix from multiple dimming brightness matrices to minimize the difference between the diffusion result and the target brightness matrix. The diffusion result is obtained by substituting the dimming brightness matrix into the diffusion formula, which is established based on the point diffusion function of each light-emitting element of the backlight module in each dimming area.

Claims

1. A decoupled control method for local dimming, executed by a computer device, comprising: Analyzing an input image, when a display device displays the input image, a target brightness is calculated for each of the plurality of dimming regions of a backlight module of the display device, thereby establishing a target brightness matrix; the target brightness matrix is ​​input into a backlight decoupling neural network model to obtain an optimal dimming brightness matrix, which contains an optimal dimming brightness for each of the dimming regions; and the optimal dimming brightness is used to control the backlight module to emit light; wherein the training dataset used in the modeling process of the backlight decoupling neural network model is related to a constrained least squares problem, the goal of which is to find the optimal dimming brightness matrix from the plurality of dimming brightness matrices such that the difference between a diffusion result and the target brightness matrix is ​​minimized, wherein the diffusion result is obtained by substituting the dimming brightness matrices into a diffusion formula, wherein the diffusion formula is established based on the point spread function (PSF) of each of the plurality of light-emitting elements of the backlight module in each of the dimming regions.

2. The decoupled control method for local dimming as described in claim 1, wherein the constrained least squares problem is expressed as follows: , where is the optimal dimming brightness matrix, is the target brightness matrix, is the diffusion formula, is one of the dimming brightness matrices, and is the diffusion result.

3. The decoupled control method for local dimming as described in claim 1, wherein the constraint condition of the constraint least squares problem is expressed as follows: , where is one of the dimming brightness matrices.

4. The decoupling control method for local dimming as described in claim 3 further includes: Analyze a plurality of historical input images to obtain a historical target brightness matrix for each of the historical input images when the display device displays the historical input images; solve the constrained least squares problem based on the historical target brightness matrix and the diffusion formula to obtain a historical optimal dimming brightness matrix corresponding to the historical target brightness matrix; and use the historical target brightness matrix and the historical optimal dimming brightness matrix of each of the historical input images to establish the training data set, and train the backlight decoupling neural network model based on the training data set.

5. The decoupled control method for local dimming as described in claim 4, wherein the interior point method (IPM) is used to solve the constrained least squares problem based on the historical target brightness matrix and the diffusion formula.

6. The decoupling control method for local dimming as described in claim 1, wherein the display device is a field sequential color (FSC) display device.

7. The decoupled control method for local dimming as described in claim 1, wherein the backlight module is direct-lit and each of the light-emitting elements is a sub-millimeter light-emitting diode (Mini LED).

8. A decoupling control method for local dimming, executed by a computer device, comprising: Brightness distribution measurements are performed on each of the plurality of dimming regions of a backlight module of a display device to obtain a point diffusion function for each of the plurality of light-emitting elements of the backlight module in each of the dimming regions. A diffusion formula is established based on the point diffusion function, and a backlight decoupling mathematical model is established based on the diffusion formula. One of a plurality of historical input images is analyzed. When the display device displays one of the historical input images, a historical target brightness is obtained for each of the dimming regions, thereby establishing a historical target brightness matrix for each of the historical input images. The historical target brightness matrix is ​​input into the backlight decoupling mathematical model to solve a constrained least squares problem. The goal is to find a historically optimal dimming brightness matrix from the plurality of historical dimming brightness matrices to minimize the difference between a diffusion result and the historical target brightness matrix, wherein the diffusion result is obtained by substituting the historical dimming brightness matrices into the diffusion formula. A training dataset is established using the historical target brightness matrix and the historical optimal dimming brightness matrix of each of the historical input images, and a backlight decoupling neural network model is trained based on the training dataset; an input image is analyzed, and when the display device displays the input image, a target brightness of each of the dimming regions is calculated to establish a target brightness matrix; the target brightness matrix is ​​input into the backlight decoupling neural network model to obtain an optimal dimming brightness matrix, which includes an optimal dimming brightness of each of the dimming regions; and the optimal dimming brightness is used to control the backlight module to emit light.

9. The decoupled control method for local dimming as described in claim 8, wherein the constrained least squares problem is expressed as follows: , where is the historical best dimming brightness matrix, is the historical target brightness matrix, is the diffusion formula, is one of the historical dimming brightness matrices, and is the diffusion result.

10. The decoupled control method for local dimming as described in claim 8, wherein the constraint condition of the constraint least squares problem is expressed as: , where is one of the historical dimming brightness matrices.

11. The decoupling control method for local dimming as described in claim 8, wherein the backlight decoupling mathematical model is solved by using the interior point method to solve the constrained least squares problem.

12. The decoupling control method for local dimming as described in claim 8, wherein the display device is a field-sequential display device.

13. The decoupled control method for local dimming as described in claim 8, wherein the backlight module is direct-lit and each of the light-emitting elements is a sub-millimeter light-emitting diode.

Images