Display device, method for establishing current correction value, and current correction system

Abstract

The present invention proposes a method for establishing a current calibration value, comprising: establishing a current setting value sequence, the current setting value sequence including a plurality of current setting values, driving a light-emitting unit, and measuring the current value of the light-emitting unit; establishing a recursive neural network, the recursive neural network including an input layer, a hidden layer, and an output layer; and inputting the plurality of current values into the hidden layer, while simultaneously inputting the plurality of current setting values sequentially into the input layer, and sequentially obtaining a plurality of calibration values from the output layer, the plurality of calibration values corresponding to the plurality of current setting values.

Description

Technical Field

[0001] The present disclosure relates to a current calibration method for establishing a current calibration value of a backlight module in a display device. Background Art

[0002] Displays are one of the most common electronic devices in modern life, used in a variety of scenarios and situations. Some displays use a backlight module to provide light. This backlight module includes multiple light-emitting diodes. Using local dimming technology, the brightness of these LEDs can be independently controlled, thereby improving the contrast of the display. However, in some conventional technologies, these LEDs are driven according to default parameters. However, due to factors such as process variation, the preset parameters may not necessarily provide light of the preset brightness. Therefore, how to calibrate these parameters is an issue of concern to those skilled in the art. Summary of the Invention

[0003] The embodiments disclosed herein propose a method for establishing a current correction value, which is applicable to a display device. The display device includes a display panel and a backlight module, the display panel includes multiple areas, the backlight module includes multiple light-emitting units, and each area corresponds to at least one light-emitting unit. The light-emitting unit is driven by the current of the backlight module to emit light to serve as the backlight source of the area of the display panel. The establishment method includes: establishing a current setting value sequence, the current setting value sequence includes multiple current setting values, driving a first light-emitting unit, and measuring the current value of the first light-emitting unit; establishing a recursive neural network, the recursive neural network includes an input layer, a hidden layer, and an output layer; and inputting the first current value into the hidden layer, while inputting multiple current setting values into the input layer in sequence, and obtaining multiple correction values from the output layer in sequence, the multiple correction values corresponding to multiple current setting values respectively.

[0004] In some embodiments, the aforementioned establishing method further includes: obtaining one of the calibration values, and driving the first light-emitting unit according to the obtained calibration value.

[0005] In some embodiments, the operation of the recurrent neural network includes the following mathematical formula.

[0006] s(d)=f2(W×f1(V×t(d)+U×m(d+1)))

[0007] Where d is the dimming level, t(d) is the current setting value, and multiple dimming levels d are set from large to small in the current setting value sequence. s(d) is the correction value. m(d+1) is the input of the hidden layer. When processing the maximum dimming level, m(d+1) is the first current value. W, V, and U are weights, and f1 and f2 are activation functions.

[0008] In some embodiments, multiple current setting values correspond to multiple driving values. The above-mentioned establishment method also includes: in a training stage, for each current setting value, driving the light-emitting unit according to the corresponding driving value, and obtaining a second current value, and using a negative feedback control method to adjust the corresponding driving value so that the second current value conforms to the corresponding current setting value, wherein the adjusted multiple driving values and the second current value corresponding to the maximum dimming level form a training sample.

[0009] In some embodiments, the activation functions f1 and f2 are sigmoid functions, linear rectifier units, or hyperbolic tangent functions. The plurality of dimming levels are an arithmetic progression.

[0010] From another perspective, an embodiment of the present disclosure proposes a display device comprising a display panel, a backlight module, and at least one circuit. The display panel comprises a plurality of regions. The backlight module comprises a plurality of light-emitting units, and each region of the display panel corresponds to at least one light-emitting unit. The light-emitting unit is driven by the current of the backlight module to emit light to serve as the backlight source of the region of the display panel. The above-mentioned circuit comprises a correction lookup table, which comprises a plurality of correction values, each of which corresponds to a plurality of dimming levels. These correction values are generated according to a recursive neural network, and the above-mentioned circuit obtains one of the correction values and generates a corrected current according to the obtained correction value to drive the light-emitting unit.

[0011] In some embodiments, the circuit includes a timing controller and a microcontroller. The timing controller calculates a driving value corresponding to the first light-emitting unit according to a local dimming algorithm. The microcontroller includes a calibration lookup table.

[0012] From another perspective, embodiments of the present disclosure provide a current calibration system comprising the aforementioned display device and an electronic terminal. The electronic terminal establishes a current calibration value according to a calibration procedure, wherein a recursive neural network operates on the electronic terminal. The calibration procedure includes: establishing a current setting value sequence, the current setting value sequence including a plurality of current setting values, driving a first light-emitting unit, and measuring the current value of the first light-emitting unit; inputting the first current value into a hidden layer of the recursive neural network, simultaneously inputting the plurality of current setting values sequentially into an input layer, and sequentially obtaining a plurality of calibration values from an output layer; and establishing a calibration lookup table corresponding to the first light-emitting unit based on the plurality of calibration values. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to make the above features and advantages of the present invention more obvious and understandable, the following detailed description is now made with reference to the accompanying drawings.

[0014] Figure 1 FIG. 5 is a schematic diagram illustrating a current calibration system according to an embodiment.

[0015] Figure 2 FIG. 1 is a schematic diagram illustrating a plurality of regions on a display panel and corresponding light-emitting units according to an embodiment.

[0016] Figure 3 FIG. 4 is a flow chart illustrating a method for generating a current correction value for training according to an embodiment.

[0017] Figure 4 is a schematic diagram illustrating a recurrent neural network according to an embodiment.

[0018] Figure 5 FIG. 4 is a flow chart illustrating a calibration procedure according to an embodiment. DETAILED DESCRIPTION

[0019] The terms “first,” “second,” etc. used herein do not particularly imply an order or sequence, but are merely used to distinguish elements or operations described with the same technical terms.

[0020] Figure 1 is a schematic diagram illustrating a current calibration system according to an embodiment. Figure 1 The current correction system 100 includes an electronic terminal 110 and a display device 120. The electronic terminal 110 can be a personal computer, a server, or various electronic devices with computing capabilities. The display device 120 includes a circuit 130, a backlight module 140, and a display panel 150. The circuit 130 includes a timing controller 131 and a microcontroller unit (MCU) 132. The microcontroller 132 can also be replaced by a field programmable gate array (FPGA), so it should not be limited to the microcontroller 132 disclosed in this embodiment. The backlight module 140 includes a plurality of light-emitting units, such as light-emitting diodes. These light-emitting diodes are driven by the current of the backlight module 140 to provide a backlight source. The display panel 150 is, for example, a liquid crystal display panel. Figure 2 is a schematic diagram illustrating multiple regions and corresponding light emitting units on a display panel according to an embodiment. Figure 1 and Figure 2 ,exist Figure 2 In the embodiment of the present invention, the display panel 150 includes 15 regions (e.g., regions 151-153), each of which corresponds to a plurality of light-emitting units (e.g., light-emitting units 141-142). Here, the brightness of each light-emitting unit can be controlled by providing different currents to increase the contrast of the image. For example, when the image to be displayed in a certain region is darker, the brightness of the corresponding light-emitting unit can be reduced. Conversely, when the image to be displayed in a certain region is brighter, the brightness of the corresponding light-emitting unit can be increased. Figure 2 This is only an example, and the present disclosure does not limit the number of regions included in the display panel 150, nor does it limit the number of light-emitting units corresponding to each region.

[0021] Please refer to Figure 1 When a picture is to be displayed, the timing controller 131 calculates the dimming level of each area on the display panel 150. This dimming level indicates how much brightness of the backlight source is required. Based on the calculated dimming level, the timing controller 131 calculates the driving value, which is used to drive the light-emitting unit to generate a specific current. In some embodiments, the driving value is positively correlated with the above-mentioned dimming level. For example, the dimming level is represented by 8 bits, and its numerical range is 0 to 255, while the driving value is represented by 10 bits, and its numerical range is 0 to 1023. The mapping relationship between the dimming level and the driving value can be linear or nonlinear, and the present disclosure is not limited to this. Due to factors such as process variation, the use of a predetermined driving value may not necessarily drive the light-emitting unit to provide the required brightness, so the driving value needs to be calibrated. The microcontroller 132 includes a plurality of correction lookup tables, each correction lookup table corresponds to a light-emitting unit, and the correction lookup table records a plurality of correction values corresponding to the driving value. The timing controller 131 can access the calibration lookup table in the microcontroller 132 based on the calculated driving value, thereby obtaining the corresponding calibration value. The timing controller 131 then drives the corresponding light-emitting unit based on the calibration value, so that the light-emitting unit can provide the desired brightness. A method for establishing a current calibration value is proposed. This method is executed by the electronic terminal 110 and utilizes a recurrent neural network (RNN) to generate the calibration value.

[0022] Specifically, Figure 3 FIG is a flow chart illustrating the generation of a training current correction value according to an embodiment. Figure 3Here, we use the light-emitting unit 141 as an example. During the training phase, the light-emitting unit 141 is first driven according to the drive value 301. Then, the current value of the light-emitting unit 141 is measured using the measurement unit 302, such as an ammeter or a power meter. In step 303, a determination is made as to whether the measured current value meets the current setting value 304. The current setting value can be an objective target value obtained by a program calculation in the electronic terminal 110, a subjective setting value requested by the user, or a specification setting value required by the manufacturer of the electronic terminal 110. This current setting value 304 corresponds to the drive value 301. For example, the drive value can be "995," and the corresponding current setting value at this drive value is 64 mA. That is, 64 mA can enable the light-emitting unit 141 to provide the brightness corresponding to the drive value "995." In step 303, it is determined whether the actual measured current value is sufficiently close to 64 mA. In some embodiments, step 303 may determine whether the measured current value is the same as the set current value 304. In some embodiments, the difference between the measured current value and the set current value 304 may also be determined to be within a certain range. If the determination result in step 303 is negative, the drive value is adjusted in step 305. For example, if the measured current value is less than the set current value 304, the initial drive value 301 may be slightly increased to form a new drive value 301, and this new drive value 301 will be used as the basis for the next drive value adjustment in step 305. If the measured current value is greater than the set current value 304, the initial drive value 301 may be slightly decreased to form a new drive value 301, and this new drive value 301 will be used as the basis for the next drive value adjustment in step 305. The light-emitting unit 141 is then driven again according to the adjusted drive value, and the above steps are repeated. If the measured current value is determined to be consistent with the set current value in step 303, the adjusted drive value is output in step 306. In other words, the negative feedback control method is used to adjust the driving value 301 so that the measured current value matches the current setting value 304 corresponding to the driving value 301. It is worth noting that the above negative feedback control method needs to be repeated for each driving value. If there are 256 dimming levels, there will also be 256 corresponding driving values. For each driving value, the negative feedback control method must be repeated to obtain an adjusted driving value. These 256 adjusted driving values are part of the training sample.

[0023] Table 1 below includes the dimming level, current setting value, original driving value, measured current value, and adjusted driving value corresponding to a certain light-emitting unit.

[0024]

[0025] Table 1

[0026] For example, referring to the second row of Table 1, when the dimming level is 255, the current setting is 64 mA, and the original drive value is 995. However, the actual measured current value is 65.3 mA, which is greater than the current setting of 64 mA. Therefore, the drive value must be reduced, and the adjusted drive value is 993. The same applies to other dimming levels. In the following description, d represents the dimming level, t(d) represents the current setting value corresponding to dimming level d, si(d) represents the original (default) drive value corresponding to dimming level d, m(d) represents the measured current value corresponding to dimming level d, and s(d) represents the adjusted drive value corresponding to dimming level d. In this embodiment, the driving value s(d) for all dimming levels is predicted based on the current value m(d) at a certain dimming level (e.g., dimming level 255"). Since measuring the current value m(d) takes a certain amount of time, if the corresponding current value m(d) is measured for each dimming level, a lot of time is required. Using the prediction method, the corrected driving value s(d) can be quickly generated.

[0027] Figure 4 is a schematic diagram illustrating a recurrent neural network according to an embodiment. Figure 4 The recurrent neural network 400 includes an input layer 410, a hidden layer 420, and an output layer 430. The input of the input layer 410 is the current setting value t(d), and the characteristic value m(d) in the hidden layer 420 is calculated using the following mathematical formula 1, where V and U are the parameters to be trained.

[0028] [Mathematical formula 1]

[0029] m(d)=V×t(d)+U×m(d+1)

[0030] The output of the output layer 430 is the adjusted driving value s(d), which is calculated using the following equation 2, where W is the parameter to be trained. Substituting equation 1 into equation 2 yields equation 3.

[0031] [Mathematical formula 2]

[0032] s(d)=W×m(d)

[0033] [Mathematical formula 3]

[0034] s(d)=W×(V×t(d)+U×m(d+1))

[0035] In addition, an activation function f1 is included between the input layer 410 and the hidden layer 420, and an activation function f2 is included between the hidden layer 420 and the output layer 430. These two activation functions can be a sigmoid function, a rectified linear unit (ReLU), or a hyperbolic tangent function (tanh), etc., but the present disclosure is not limited to this. Since these two activation functions are not general arithmetic functions, the two activation functions are mainly mapped to a certain interval, and the function changes in this interval are quite smooth, and do not produce positive infinity or negative infinity like the arithmetic function. This helps to stabilize the correction of the driving value s(d). Under this premise, after adding the activation function, the above mathematical formula 3 can be rewritten as the following mathematical formula 4.

[0036] [Formula 4]

[0037] s(d)=f2(W×f1(V×t(d)+U×m(d+1)))

[0038] Here, the input current setting values t(255)~t(0) are a sequence, and the driving value s(d) output by the recurrent neural network 400 is also a sequence. Generally speaking, the recurrent neural network 400 can be represented in an expanded manner. For example, when the input is t(d), the output is s(d), and so on. It is worth noting that here, the dimming level d is an arithmetic progression sorted from large to small, that is, the current setting values t(255)~t(0) are sequentially input to the input layer 410, and the eigenvalue m(d) is calculated based on the eigenvalue m(d+1) of the previous iteration. When processing the highest dimming level (d=255), the eigenvalue m(256) is the current value actually measured after driving the light-emitting unit based on the original driving value si(255), for example, 65.3 mA in the second row of Table 1 above. In other words, only m(256) needs to be actually measured, and the other m(255)~(0) are all calculated by the above mathematical formula 1. The training sample requires at least the current value m(256) and all adjusted driving values s(255) to s(0).

[0039] Here, each light-emitting unit can provide a training sample. After collecting multiple training samples, the recursive neural network 400 can be trained accordingly. The trained recursive neural network 400 can be used to predict the corrected driving value. Specifically, Figure 5 FIG2 is a flow chart illustrating a calibration procedure according to an embodiment of the present invention. The calibration procedure is executed by the electronic terminal 110 to establish a calibration lookup table for a certain light emitting unit. Figure 5Here, taking the light-emitting unit 142 as an example, a current setting value sequence 501 is first established. The current setting value sequence includes current setting values t(255) to t(0), and the dimming level is an arithmetic progression arranged in sequence from "255" to "0". Next, a driving value 502 corresponding to one of the current setting values is obtained. This embodiment uses the driving value si(255) corresponding to the maximum dimming level. The light-emitting unit 142 is driven according to the obtained driving value 502, and then the current value 503 of the light-emitting unit 142 is measured using the measurement unit 302. Then, the current value 503 is input into the hidden layer of the recurrent neural network 400 as m(d+1), and the current setting values t(255) to t(0) are input into the input layer in sequence. Then, the correction values 504 are obtained from the output layer in sequence. These correction values 504 can be used to establish a correction lookup table. For example, the correction lookup table records the correction values corresponding to the dimming values 255 to 0 in sequence ("999", "989" ..., "3", "0"). This correction lookup table can be stored in the microcontroller 132. When the dimming value is "254", the correction value "989" can be found through this correction lookup table. Then the backlight module 140 can generate a corrected current based on this correction value to drive the light-emitting unit 142 to emit light, so as to serve as the backlight source of at least one area in the display panel. In other words, since the output current of each light-emitting area has been corrected, a uniform brightness characteristic can be obtained to avoid the occurrence of uneven brightness in the light-emitting area. Combined with the existing regional dimming technology, it can ensure that each light-emitting area after regional dimming can achieve the desired partition brightness.

[0040] exist Figure 5 In the process described above, only the current value 503 needs to be measured once to obtain all the calibration values. In contrast, the conventional technology requires measuring the current value corresponding to each driving value and then calculating the adjusted calibration value through negative feedback control. Therefore, generating calibration values using the method of the above embodiment can save a lot of time.

[0041] In the above embodiment, the driving value 502 uses the driving value si(255) corresponding to the maximum dimming level, but in other embodiments, the driving value corresponding to any dimming level can also be used, and the corresponding current value needs to be used in the training stage. For example, if the measured m(128) is input to the hidden layer, the current setting value of the input layer can be t(127)~t(0), (255)~t(128), that is, the current setting value t(127) must be input first. Ordinary technicians in this field should be able to design different current setting value sequences based on the above content, and the present disclosure is not limited to the above embodiment.

[0042] Although the present invention has been disclosed above through embodiments, the present invention is not limited thereto. Any person skilled in the art should be able to make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be based on the scope defined by the appended claims.

[0043] Reference Signs List

[0044] 100: Current correction system

[0045] 110: Electronic Terminal

[0046] 120: Display device

[0047] 130: Circuit

[0048] 131: Timing Controller

[0049] 132: Microcontroller

[0050] 140: Backlight module

[0051] 141, 142: Light-emitting unit

[0052] 150: Display panel

[0053] 151, 152: Area

[0054] 301: driving value

[0055] 302: Measurement unit

[0056] 303, 305, 306: Steps

[0057] 304: Current setting value

[0058] 400: Recurrent Neural Networks

[0059] 410: Input layer

[0060] 420: Hidden layer

[0061] 430: Output layer

[0062] 501: Current set value sequence

[0063] 502: Driver value

[0064] 503, 610: Current value

[0065] 504: Calibration value

[0066] 620: Lookup table.

Claims

1. A method for establishing a current correction value, applicable to a display device, the display device comprising a display panel and a backlight module, the display panel comprising a plurality of regions, the backlight module comprising a plurality of light-emitting units, each of the plurality of regions corresponding to at least one of the plurality of light-emitting units, the plurality of light-emitting units being driven by current from the backlight module to emit light, thereby serving as backlight sources for the plurality of regions of the display panel, the method comprising: establishing a current setting value sequence, wherein the current setting value sequence includes a plurality of current setting values, driving a first light-emitting unit among the plurality of light-emitting units, and measuring a first current value of the first light-emitting unit; Establishing a recursive neural network, wherein the recursive neural network includes an input layer, a hidden layer, and an output layer; as well as The first current value is input into the hidden layer, the multiple current setting values are input into the input layer in sequence, and multiple correction values are obtained from the output layer in sequence, the multiple correction values respectively corresponding to the multiple current setting values, The operation of the recursive neural network includes the following mathematical formula: Wherein, d is one of a plurality of dimming levels, t(d) is one of the plurality of current setting values, the plurality of dimming levels d are set from large to small in the current setting value sequence, s(d) is one of the plurality of correction values, m(d+1) is the input of the hidden layer, when one of the plurality of dimming levels is the maximum dimming level, m(d+1) is the first current value, W, V and U are weights, and f1 and f2 are activation functions.

2. The method for establishing a current correction value according to claim 1 , further comprising: One of the plurality of correction values is obtained, and the first light emitting unit is driven according to the obtained correction value.

3. The method for establishing a current correction value according to claim 1, wherein: The plurality of current setting values correspond to a plurality of driving values, and the establishing method further comprises: During the training phase, for each current setting value, one of the plurality of light-emitting units is driven according to the corresponding driving value to obtain a second current value. The corresponding driving value is then adjusted using a negative feedback control method so that the second current value conforms to the corresponding current setting value. The adjusted plurality of driving values and the second current value corresponding to the maximum dimming level form a training sample.

4. The method for establishing a current correction value according to claim 1, wherein: The activation functions f1 and f2 are sigmoid functions, linear rectification units or hyperbolic tangent functions, and the multiple dimming levels are an arithmetic progression.

5. A display device comprising: a display panel comprising a plurality of regions; a backlight module comprising a plurality of light-emitting units, wherein each of the regions of the display panel corresponds to at least one of the plurality of light-emitting units, and the plurality of light-emitting units are driven by current from the backlight module to emit light, thereby serving as backlight sources for the plurality of regions of the display panel; as well as At least one circuit includes a correction lookup table, the correction lookup table includes a plurality of correction values, the plurality of correction values respectively corresponding to a plurality of dimming levels, the plurality of correction values are generated according to the current correction value establishment method described in any one of claims 1 to 4, the at least one circuit obtains one of the plurality of correction values and generates a corrected current according to the obtained correction value to drive a first light-emitting unit among the plurality of light-emitting units. The display device according to claim 5 , wherein: The at least one circuit includes a timing controller and a microcontroller. The timing controller calculates a driving value corresponding to the first light-emitting unit according to a local dimming algorithm. The microcontroller includes the correction lookup table.

7. A current correction system for a display device, comprising the display device according to claim 5 or 6 and an electronic terminal, wherein the electronic terminal establishes a current correction value according to a correction program, wherein the recursive neural network runs on the electronic terminal and includes an input layer, a hidden layer, and an output layer, and the correction program includes: establishing a current setting value sequence, wherein the current setting value sequence includes a plurality of current setting values, driving the first light-emitting unit, and measuring a first current value of the first light-emitting unit; Inputting the first current value into the hidden layer, inputting the plurality of current setting values into the input layer in sequence, and obtaining the plurality of correction values from the output layer in sequence; as well as The correction lookup table corresponding to the first light emitting unit is established according to the plurality of correction values.

8. The current correction system for a display device according to claim 7, wherein: The activation functions f1 and f2 are sigmoid functions, linear rectification units or hyperbolic tangent functions, and the multiple dimming levels are an arithmetic progression.

Images