KR20210047086A - Electronic device and method for analyzing power comsumption for display panel thereof - Google Patents

Abstract

According to an aspect according to the technical idea of the present disclosure, provided is a method for analyzing power consumption of a display panel, which comprises the steps of: performing downsampling on at least one image included in image data displayed on a display panel; generating a power consumption model of the display panel in consideration of the current dependence between sub-pixels included in the display panel; and calculating power consumption of the display panel based on at least one downsampled image by using the power consumption model of the display panel.

Description

Electronic device and its display panel power consumption analysis method {ELECTRONIC DEVICE AND METHOD FOR ANALYZING POWER COMSUMPTION FOR DISPLAY PANEL THEREOF}

The technical idea of the present disclosure relates to an electronic device and a method of analyzing power consumption of a display panel thereof, and more particularly, an electronic device capable of analyzing the power consumption of a display panel in real time while maintaining high accuracy, and its It relates to a method for analyzing power consumption of a display panel.

Mobile devices equipped with AMOLED (Active Matrix Organic Light Emitting Diode) displays consume a lot of power as they are used for Internet, gaming, etc., but the need for power savings is increasing due to the limitation of using only a set battery capacity. . In particular, as the proportion of AMOLED displays in the power consumption of mobile devices is increasing, there is a growing need for a power model that can accurately predict and manage the power consumption of AMOLED displays in real time inside the mobile device.

In the AMOLED display panel, the pixel configuration of the displayed content affects the power consumption. In particular, since R, G, and B sub-pixels in AMOLED displays use a common power source, the dependence between sub-pixels must be taken into account to accurately estimate power. Therefore, a power consumption analysis model (hereinafter referred to as a power model) of various AMOLED display panels in consideration of dependence between sub-pixels has been studied.

Among the various power models, the Hong model (Hong et al., 3 Channel Dependency-Based Power Model for Mobile AMOLED Displays) is a model with improved accuracy taking into account all dependencies existing between R, G, and B 3 channels. The Hong model has the highest accuracy among the existing power models, and is consumed by the AMOLED display panel based on the power consumption that appears when outputting a black screen and the power consumption when outputting images with changed R, G, and B pixels respectively. The power was analyzed. The Hong model can be described through Equations 1 and 2 as follows.

Here, i may mean an arbitrary pixel in the display panel, and n may mean the total number of pixels in the display panel. f(R _i ), h(G _i ) and k(B _i ) are functions of current consumption of the R, G and B subpixels, respectively, and C _Ipanel is the current consumption of the panel measured when it is static, that is, on a black screen. Means. D ₂ (R _{i ,} G _{i ,} B _i ) is a 2-channel dependent current compensation term that considers more accurate dependence between two channels, RG, RB, and GB, and D ₃ (R _{i ,} G _{i ,} B _i ) is RGB This is a 3-channel dependent current compensation term that considers the dependence of the liver.

Using the Hong model, the power consumed by the AMOLED display panel can be estimated with high accuracy. Using this power model, it can be widely used, such as performing a dynamic voltage scaling technique (Dynamic Voltage Scaling, DVS), a method to reduce the power consumption of an AMOLED display, or predicting the remaining battery capacity by using it in the calculation of an actual battery model. have.

However, the Hong model is a model that considers the dependence of all three channels of R, G, and B, and shows high accuracy, but due to the high computational complexity, it shows a slow computation time of 2.96 seconds per image on a mobile device. Therefore, using the Hong model, there is a limitation that it is not possible to analyze power consumption in real time on a mobile device. Therefore, it can be said that a new method is needed to enable various applications such as DVS and battery remaining capacity prediction by analyzing power consumption in real time.

An electronic device and a method for analyzing power consumption of a display panel thereof according to the technical idea of the present disclosure are to be able to analyze power consumption of a display unit in real time using a power model while maintaining high accuracy.

The technical problem to be achieved by the method of analyzing power consumption of an electronic device and its display panel according to the technical idea of the present disclosure is not limited to the above-mentioned problem, and another problem not mentioned is clearly understood by those skilled in the art from the following description. It will be possible.

According to an aspect of the inventive concept, there is provided a method of analyzing power consumption of a display panel, the method comprising: performing downsampling on at least one image included in image data displayed on the display panel; Generating a power consumption model of the display panel in consideration of a current dependence relationship between sub-pixels included in the display panel; And calculating power consumption of the display panel based on the downsampled at least one image using the power consumption model of the display panel.

According to an exemplary embodiment, the power consumption model of the display panel is expressed as a product of a current of the display panel and a voltage of the display panel, and the current of the display panel is Current-hereinafter referred to as'channel independent current'-is subtracted from a current according to a dependence relationship between two sub-pixels among the first to third sub-pixels-hereinafter referred to as a '2-channel current dependence relationship', and It can be expressed as adding a current according to a dependency relationship between the first to third sub-pixels-hereinafter referred to as a'three-channel current dependency relationship'-and adding a fixed current of the display panel.

According to an exemplary embodiment, the two-channel current dependence relationship may be expressed as a first polynomial for any one of the two subpixels, and the first polynomial is a polynomial less than 5th order. I can.

According to an exemplary embodiment, the maximum order of the first polynomial may be 3rd order.

According to an exemplary embodiment, the three-channel current dependence relationship may be expressed as a second polynomial for any one of the first to third subpixels, and the second polynomial is less than 5th order. It can be a polynomial.

According to an exemplary embodiment, the maximum order of the second polynomial may be third.

According to an exemplary embodiment, the calculating of the power consumption of the display panel is dependent on the channel independent current and the two-channel current for at least two or more pixels among a plurality of pixels of the downsampled at least one image. Any one of the relationship and the three-channel current dependence relationship can be calculated in parallel.

According to an exemplary embodiment, the calculating of the power consumption of the display panel includes the channel independent current, the 2-channel current dependency relationship, and the 3-channel for each of a plurality of pixels of the downsampled at least one image. At least two of the current dependence relationships can be divided and calculated in parallel.

According to an exemplary embodiment, performing downsampling on the at least one image may include downsampling by grouping adjacent pixels of the at least one image.

According to an exemplary embodiment, the step of grouping and downsampling adjacent pixels of the at least one image includes calculating an average value of adjacent pixels in each group, or selecting and downsampling a representative pixel among adjacent pixels in each group. I can.

According to an exemplary embodiment, the calculating of power consumption of the display panel includes: dividing the downsampled at least one image into a predetermined number of regions; Determining index values of a lookup table based on an average value of subpixels of each of the divided regions; Accessing the lookup table based on index values of the lookup table to obtain power consumption for the partitioned regions; And calculating power consumption of the display panel based on power consumption of the divided regions.

According to an exemplary embodiment, the lookup table may store a value obtained by pre-calculating the power consumption of the divided regions based on the power consumption model of the display panel.

According to an exemplary embodiment, the determining of the index values of the lookup table may include: hashing by dividing an average value of subpixels of each of the divided regions by a predetermined integer; And determining index values of the lookup table based on the hashing result.

According to another aspect according to the technical idea of the present disclosure, a display panel; And a control unit for controlling the display panel, wherein the control unit comprises: a downsampling unit for downsampling at least one image included in image data displayed on the display panel; A power consumption model generator configured to generate a power consumption model of the display panel in consideration of a current dependence relationship between subpixels included in the display panel; And a power consumption calculator configured to calculate power consumption of the display panel based on the down-sampled at least one image using the power consumption model of the display panel.

In the method of analyzing power consumption of an electronic device and its display panel according to embodiments of the inventive concept, while maintaining high accuracy, the power consumption of the display panel can be analyzed in real time using a power model, and based on this Thus, there is an effect that can be used to predict the remaining battery capacity by applying a dynamic voltage variable technique, which is a method for reducing power consumption, or by using it in calculating an actual battery model.

The effects that can be obtained by the method for analyzing power consumption of an electronic device and its display panel according to the technical idea of the present disclosure are not limited to the above-mentioned effects, and other effects not mentioned are the technology to which the present disclosure belongs from the following description. It will be clearly understood by those of ordinary skill in the field.

In order to more fully understand the drawings cited in the present disclosure, a brief description of each drawing is provided.
1 is a diagram schematically illustrating a partial configuration of an electronic device according to an embodiment of the inventive concept.
2 is a diagram schematically illustrating a partial configuration of a control unit according to an embodiment according to the technical idea of the present disclosure.
3 is a flowchart illustrating a method of analyzing power consumption of a display unit of an electronic device according to an exemplary embodiment of the inventive concept.
4A is a table showing an embodiment of an algorithm of a power model for implementing power consumption analysis of a display unit using an image downsampling and lookup table according to an embodiment of the inventive concept.
FIG. 4B is a diagram showing an embodiment of an algorithm of a power model for implementing an image downsampling, analysis of power consumption of a display unit using a plurality of threads and NEON instructions according to an embodiment of the inventive concept of the present disclosure to be.
FIG. 5 is a diagram for describing a method for analyzing power consumption of a display unit based on parallelization of a command level, according to an embodiment of the inventive concept.
FIG. 6 is a diagram illustrating a method for analyzing power consumption of a display unit based on parallelization of a thread level, according to an embodiment according to the technical idea of the present disclosure.
7A is an analysis of power consumption of the display unit by limiting the polynomial order of the 2-channel dependent current compensation term and the 3-channel dependent current compensation term of the power model in order to minimize computational complexity according to an embodiment of the inventive concept. It is a figure for demonstrating the method for doing.
7B is an analysis of power consumption of the display unit by limiting the polynomial order of the 2-channel dependent current compensation term and the 3-channel dependent current compensation term of the power model in order to minimize computational complexity according to an embodiment of the inventive concept. It is a diagram for explaining the effect of the following method.
FIG. 8 is a diagram for explaining a method for analyzing power consumption of a display unit by using a lookup table with divided images according to an embodiment of the inventive concept.
9 is a diagram illustrating a software platform for implementing a method of analyzing power consumption of a display unit of an electronic device according to an embodiment of the inventive concept.
10 is a table comparing effects of a method for analyzing power consumption of a display unit of an electronic device based on a power model according to various embodiments of the present disclosure.
11 is a graph comparing average execution times of a method for analyzing power consumption of a display unit of an electronic device based on various settings of a power model according to an exemplary embodiment of the inventive concept.
12 is a graph comparing an average error rate of a method of analyzing power consumption of a display unit of an electronic device based on various settings of a power model according to an exemplary embodiment of the inventive concept.

The technical idea of the present disclosure is that various changes may be made and various embodiments may be provided, and specific embodiments will be illustrated in the drawings and described in detail. However, this is not intended to limit the technical idea of the present disclosure to a specific embodiment, it should be understood to include all changes, equivalents, and substitutes included in the scope of the technical idea of the present disclosure.

In describing the technical idea of the present disclosure, if it is determined that a detailed description of a related known technology may unnecessarily obscure the subject matter of the present disclosure, a detailed description thereof will be omitted. In addition, numbers (eg, first, second, etc.) used in the description process of the present disclosure are merely identification symbols for distinguishing one component from another component.

In addition, in the present disclosure, when one component is referred to as "connected" or "connected" with another component, the one component may be directly connected or directly connected to the other component, but specifically It should be understood that as long as there is no opposite substrate, it may be connected or may be connected via another component in the middle.

In addition, terms such as "~ unit", "~ group", "~ character", and "~ module" described in the present disclosure mean a unit that processes at least one function or operation, which is a processor or microcomputer. Processor (Micro Processer), Micro Controller, CPU (Central Processing Unit), GPU (Graphics Processing Unit), APU (Accelerate Processor Unit), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA It can be implemented by hardware or software such as (Field Programmable Gate Array), or a combination of hardware and software.

In addition, it is intended to clarify that the division of the constituent parts in the present disclosure is merely divided by the main functions that each constituent part is responsible for. That is, two or more constituent parts to be described below may be combined into one constituent part, or one constituent part may be divided into two or more for each more subdivided function. In addition, each of the constituent units to be described below may additionally perform some or all of the functions of other constituent units in addition to its own main function, and some of the main functions of each constituent unit are different. It goes without saying that it can also be performed exclusively by.

Hereinafter, embodiments according to the technical idea of the present disclosure will be described in detail in order.

1 is a diagram schematically illustrating a partial configuration of an electronic device according to an embodiment of the inventive concept.

The electronic device 100 may include a communication unit 110, an output unit 120, a user input unit 130, a power supply unit 140, a memory 150, and an interface unit 160. The electronic device 100 illustrated in FIG. 1 may be a mobile device such as a smart phone or a tablet PC, but is not limited thereto. In addition, since the components shown in FIG. 1 are not essential, the electronic device 100 may be implemented as a device having more components or fewer components. Hereinafter, the above components will be described in order.

The communication unit 110 may include one or more units that enable communication between the electronic device 100 and other devices. For example, the communication unit 110 may include a receiver 112, a transmitter 114, and the like. The receiver 112 receives a signal transmitted from the other device through a predetermined communication channel. Here, signals transmitted from other devices may include signals having terminal information of other devices and signals having various data. The transmitter 114 may transmit a signal including various types of information from the electronic device 100 to the outside, and may transmit a signal with command information so that another device performs a specific function.

The output unit 120 is for generating output related to visual and audio, and may include a display unit 122 and an audio output unit (not shown).

The display unit 122 may include a display module. The display module may include a display panel and a display driver. The display driver may drive a pixel, and may be configured with an analog circuit or a digital circuit. The display unit 122 outputs various types of information processed by the electronic device 100 as an image, a moving image, or a text. The display unit 122 may be formed of, for example, an AMOLED display. Here, the AMOLED display is a concept including a flat and flexible type, and includes an OLED (Organic Light-Emitting Diode) display in the form of an improved AMOLED display similar to the AMOLED display.

The user input unit 130 generates input data for a user to operate a device. The user input unit 130 may be composed of a key pad, a dome switch, a touch pad (positive pressure/electrostatic), a jog wheel, a jog switch, and the like.

The power supply unit 140 supplies power necessary for operation to each of the components based on external power and power applied from an internal battery (not shown) under the control of the control unit 170.

The memory 150 may store at least one program for operation of the control unit 170 or the like, and may temporarily store input/output data. For example, the memory 150 may store a lookup table to be described later, at least one program for performing a method of analyzing power consumption of the display unit of the electronic device 100, and the like.

The memory 150 is a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (eg, SD or XD memory, etc.), and RAM. At least one of (Random Access Memory, RAM), SRAM (Static Random Access Memory), ROM (ReadOnly Memory, ROM), EEPROM (Electrically Erasable Programmable ReadOnly Memory), PROM (Programmable ReadOnly Memory) magnetic memory, magnetic disk, optical disk It may be a type of storage medium.

The interface unit 160 functions as a means for communicative connection with all external devices connected to the electronic device 100.

The controller 170 generally controls the overall operation of the electronic device 100 and may perform a display unit power consumption analysis method, which will be described later. In addition, the controller 170 may control the output of an image through the display unit 122 based on a power consumption value of the display unit determined according to a result of the method for analyzing power consumption of the display unit. The control unit 170 will be described in more detail below with reference to FIG. 2.

2 is a diagram schematically illustrating a partial configuration of a control unit according to an embodiment according to the technical idea of the present disclosure.

Referring to FIG. 2, the control unit 170 includes a downsampling unit 171, a power consumption model generation unit 172, a power consumption calculation unit 173, a region division unit 174, and a lookup table generation unit 175. Can include.

Since the components shown in FIG. 2 are not essential, the control unit 170 may have more components or be implemented with fewer components.

The downsampling unit 171 may perform downsampling on at least one image included in image data displayed on the display unit 122 (hereinafter, referred to as a display panel).

For example, the downsampling unit 171 may downsample by grouping adjacent pixels of at least one image included in the image data. The downsampling unit 171 may determine a plurality of pixel groups by dividing the image into a plurality of pixel groups and grouping adjacent pixels in the image into a plurality of pixel groups. Here, the pixel groups may each include the same number of pixels and may have the same size. For example, the size of the pixel groups may be MxN (M and N are integers).

The downsampling unit 171 may sample R (Red), G (Green), and B (Blue) data by calculating an average value of pixels included in the pixel group. Alternatively, the downsampling unit 171 may select a representative pixel from among pixels included in the pixel group and sample R (Red), G (Green), and B (Blue) data based on the selected pixel.

Meanwhile, the downsampling may be 1/4 or 1/8 downsampling. Here, 1/4 downsampling is a process for reducing the size of the input image to 1/4. In this case, the pixel group may include 4 pixels and may be a 2x2 block. 1/8 downsampling is a process for reducing the size of an input image to 1/8, in this case, a pixel group may include 8 pixels, and may be a 4x2 or 2x4 block.

The downsampling unit 171 may perform downsampling on at least one image and output at least one downsampled image. In this case, the downsampled at least one image may be an image having a size smaller than the size of the input image. For example, the image before downsampling may be an HD (720x1280) level image, and the downsampled image may be a 360x640 image.

The power consumption model generator 172 may generate a power consumption model of the display panel in consideration of a current dependence relationship between subpixels included in the display panel of the display unit. In this case, the power consumption model of the display panel is expressed as a product of the voltage and current of the display panel.

The power consumption model generator 172 may generate a power consumption model of the display panel in consideration of a current dependence relationship between two types of subpixels and a current dependence relationship between three types of subpixels. Here, the sub-pixel may include three types of sub-pixels (R, G, and B sub-pixels).

The current of the display panel constituting the power consumption model of the display panel may be expressed by Equation 3 below.

At this time, f(R _i ), h(G _i ), k(B _i ) denote current consumption according to the values of R, G, and B subpixels, and f(R _i ), h(G _i ), k (B _i ) is. It can be expressed as a function of current consumption of each of the R, G and B sub-pixels. C _panel means the fixed current of the display panel.

D ₂ (Ri,G _i ,B _i ) represents the current dependence relationship (i.e., 2-channel current dependence relationship) between two types of sub-pixels, which is a 2-channel-dependent current compensation term, expressed as a 2-channel-dependent current compensation function. Can be. D ₃ (Ri,G _i ,B _i ) represents the current dependence relationship (i.e., 3-channel current dependence relationship) between three types of sub-pixels, which is a 3-channel-dependent current compensation term, expressed as a 3-channel-dependent current compensation function. Can be.

That is, the current of the display panel is calculated by subtracting the current according to the 2-channel current dependence relationship from the current of each of the three types of sub-pixels as shown in Equation 3, and adding the current according to the 3-channel current dependence relationship, and the display It can be expressed by adding the fixed current of the panel.

The two-channel current dependence relationship can be expressed by Equation 4 below.

In this case, α(R _i ,G _i ) represents the dependent current relationship between the R sub-pixel and the G sub-pixel, and is a dependent current compensation term between the R sub-pixel and G sub-pixel, which compensates for the dependent current between the R sub-pixel and G sub-pixel. It can be expressed as a function. β(R _i ,B _i ) represents the dependent current relationship between R sub-pixel and B sub-pixel, as a dependent current compensation term between R sub-pixel and B sub-pixel, as a dependent current compensation function between R sub-pixel and B sub-pixel. Can be expressed. γ(G _i ,B _i ) represents the dependent current relationship between G sub-pixel and B sub-pixel, as a dependent current compensation term between G sub-pixel and B sub-pixel, as a dependent current compensation function between G sub-pixel and B sub-pixel. Can be expressed.

That is, the current according to the two-channel current dependence relationship is a current according to the dependence relationship between the first sub-pixel and the second sub-pixel, and the current according to the dependence relationship between the first sub-pixel and the third sub-pixel according to Equation 4 above. , May be expressed as a sum of currents according to the dependency relationship between the second sub-pixel and the third sub-pixel.

Current according to the dependency relationship between the three sub-pixel pairs is expressed as Equations 5 to 7 below.

Here, a _k , b _k , c _k denotes a coefficient of a polynomial, and its value can be expressed as a function for one sub-pixel. For example, a _k (R _i ) may have a different value depending on the value of the R subpixel. m is a measure of the accuracy of power modeling, and may have, for example, one of an integer of 2 to 5.

The three-channel dependent current relationship can be expressed by Equation 8 below.

That is, D ₃ (Ri,G _i ,B _i ) represents the current dependence relationship (3-channel current dependence relationship) between three types of sub-pixels, which is a 3-channel-dependent current compensation term, expressed as a 3-channel-dependent current compensation function. Can be. Here, m is a measure of the accuracy of power modeling, and may have a value of 2 to 5. z _k (R _i ,B _i ) represents trend information on the graph of the current value according to the change of the dependent R sub-pixel and B sub-pixel in the state where the G sub-pixel is fixed. It can be expressed as a function for Specifically, z _k (R _i ,B _i ) is expressed as in Equation 9 below.

Here, x _u (Bi) denotes trend information on a graph of a current value according to a sub-pixel B in a state in which the R sub-pixel is fixed, and may be expressed as a function of the B sub-pixel. p is a measure of the accuracy of the power model, and may have a value of 2 to 5.

In order to generate a power consumption model, the power consumption model generator 172 subtracts the current of the display panel from (i) the current for each of the three types of subpixels (ii) the current according to the two-channel dependency relationship, and , (iii) It can be expressed by adding the current according to the 3-channel dependence relationship, and (iv) adding the fixed current of the display panel.

In order to generate a power consumption model, the power consumption model generation unit 172 generates a current according to a two-channel dependency relationship: (i) a current according to a dependency relationship between the first sub-pixel and the second sub-pixel, and (ii) the first sub-pixel. It can be expressed as the sum of the current according to the dependence relationship between the pixel and the third subpixel, and (iii) the current according to the dependence relationship between the second subpixel and the third subpixel.

The power consumption model generation unit 172 generates the current according to the two-channel dependency relationship (i), (ii), and (iii) of each sub-pixel pair (a first sub-pixel, a second sub-pixel, and a first sub-pixel). The third sub-pixel, the second sub-pixel, and the third sub-pixel) may be expressed as a polynomial according to any one type of sub-pixel. In this case, the power consumption model generation unit 172 may express the polynomial as a polynomial of less than 5th order. For example, the power consumption model generator 172 may express the maximum order of the polynomial as a third order.

The power consumption model generator 172 may express the current according to the three-channel dependency relationship as a polynomial for the third sub-pixel, based on the dependency relationship between the first sub-pixel, the second sub-pixel, and the third sub-pixel. The coefficients for each order of the, polynomial may be expressed as a polynomial for one of the first sub-pixel and the second sub-pixel based on a dependency relationship between the first sub-pixel and the second sub-pixel.

The power consumption model generator 172 may express the polynomials as polynomials of less than 5th order. For example, the power consumption model generator 172 may express the maximum order of the polynomials as a third order.

That is, the power consumption model generation unit 172 may simplify the complexity of calculating the power consumption by limiting the maximum order of a polynomial of a current according to a two-channel dependence relationship or a current according to a three-channel dependence relationship.

The power consumption calculator 173 may calculate the power consumption of the display panel for the downsampled image based on the generated power consumption model of the display panel.

According to an embodiment, the power consumption calculation unit 173 performs a calculation operation on a specific term (one of the terms of Equations 3 to 9) with respect to the pixels in the image repeatedly based on the power model through command level parallelization. When executing, it is possible to perform calculations on a specific term for several pixels with a single command.

For example, the power consumption calculation unit 173 uses a single NEON instruction based on a SIMD (Single Instruction Multiple Data) extended architecture to provide a first sub-pixel and a second sub-pixel corresponding to each of the plurality of pixels. Current according to the dependency relationship between (e.g., β(R _i ,B _i ) in Equation 4 is calculated in parallel; β(R _i ,B _i ), β(R _i+1 ,B _i+2 ) Calculate β(R _{i + x} ,B _{i +x} ), where x represents the loop unrolling degree of the NEON instruction).

Although NEON instruction is one instruction, it can process multiple data under SIMD (Single Instruction Multiple Data) extended architecture. Therefore, when the NEON instruction is used to repeatedly calculate a specific term for pixels in an image based on the power model, calculations for a specific term may be performed in parallel for several pixels.

According to another embodiment, the power consumption calculation unit 173 performs a task (calculation) on several terms (terms of Equations 3 to 9) to a plurality of threads based on a power model through thread level parallelism. By distributing, the intermediate calculation result value is generated, and finally, the power consumption can be calculated by summing the intermediate calculation result values in each thread.

For example, the power consumption calculation unit 173 uses a POSIX thread to each of a plurality of threads: (i) Current for each of the three types of sub-pixels (ii) Dependence between the first sub-pixel and the second sub-pixel. The calculation work on the current according to the relationship, (iii) the current according to the dependence relationship between the first sub-pixel and the third sub-pixel, and (iv) the current according to the dependence relationship between the second and third sub-pixels is distributed. You can, and you can perform computations in each thread in parallel.

Meanwhile, the power consumption calculator 173 may calculate the power consumption of the display panel for the downsampled image using a pre-stored lookup table.

The lookup table may be a table in which pre-calculated values of power consumption corresponding to an average value of sub-pixels in an image region based on a power consumption model of the display panel generated by the power consumption model generator 172 are stored. . The lookup table may be generated in advance by the lookup table generation unit 175, but is not limited thereto.

The lookup table may be LUTX (X is an integer), where X is a range of 0 to 255 (a range when the bit depth is 8, and the value changes when the bit depth is different). It may mean a range of sub-pixel values included in the unit for storing power consumption. That is, in the case of LUT5, power consumption corresponding to the average value of the subpixels may be stored in LUT5 for every 5 subpixel value range of the subpixel value range of 0 to 255.

For example, if a sub-pixel has three types (R, G, B), the lookup table may be a three-dimensional table.

The power consumption calculation unit 173 may receive a divided image from the region dividing unit 174. The region dividing unit 174 may divide at least one image into NxM (M and N are integers) regions. In this case, all of the regions may have the same size. For example, the region dividing unit 174 may divide at least one downsampled image into NxM (M and N are integers) regions.

The power consumption calculator 173 may determine the index value of the lookup table by hashing the average value of the subpixels of each region divided by the region dividing unit 174. The lookup table may be accessed through an index of a 3D array, and the power consumption calculator 173 may determine the index for each divided area. For example, if the average value of the R, G, B subpixels in the divided area is 78,35,67 and the lookup table is LUT30, the average value of the R, G, B subpixels is hashed to correspond to each subpixel. Indexes 2, 1, and 2 may be determined.

The power consumption calculation unit 173 may access the lookup table based on the index value of the lookup table to obtain power consumption for each region divided by the image division unit 174, and add the power consumption for each region. The power consumption of the display panel can be calculated.

3 is a flowchart illustrating a method of analyzing power consumption of a display unit of an electronic device according to an exemplary embodiment of the inventive concept.

In step S305, the electronic device 100 may perform downsampling on at least one image included in image data displayed on the display panel.

In operation S310, the electronic device 100 may generate a power consumption model of the display panel in consideration of a current dependence relationship between subpixels included in the display panel. For example, the electronic device 100 may generate a power consumption model of the display panel in consideration of a current dependence relationship between two types of subpixels and a current dependence relationship between three types of subpixels.

In operation S315, the electronic device 100 may calculate power consumption of the display panel based on at least one down-sampled image based on the power consumption model of the display panel.

4A is a table showing an embodiment of an algorithm of a power model for implementing analysis of power consumption of a display unit using an image downsampling and lookup table according to an embodiment of the inventive concept.

Referring to FIG. 4A, an algorithm of a power model (optimum power model A) for outputting an estimated power consumption of a display unit corresponding to the sub-pixel value of an image is shown. Optimal power model A is an optimized model while minimizing computational complexity, and is a model that combines 1/4 or 1/8 image (image) downsampling, lowering the order of the equation, and utilizing a lookup table.

The controller 170 may perform 1/4 or 1/8 image downsampling. The control unit 170 divides the image to obtain NxM divided regions (sub_images), and the subpixels of each sub-image (sub_image) (by executing a loop statement as many as the number of divided regions, and performing the operation of each divided region) By hashing the average value, the LUT indexes (R_idx, G_idx, B_idx) may be obtained. The controller 170 accesses the value of the corresponding lookup table (LUT[R_idx][G_idx][B_idx]) based on the indexes (R_idx, G_idx, B_idx) of the LUT, and The values LUT[R_idx][G_idx][B_idx]) are summed to determine the expected power consumption of the display unit, and the estimated power consumption may be output.

In the case of the optimal power model A, the amount of calculation required for calculating the estimated power consumption can be reduced to the maximum by using a look-up table, including image downsampling and order limitation, so that the estimated power consumption is reduced based on the corresponding power model inside the electronic device. It can be calculated quickly.

FIG. 4B is a diagram showing an embodiment of an algorithm of a power model for implementing an image downsampling, analysis of power consumption of a display unit using a plurality of threads and NEON instructions according to an embodiment of the inventive concept of the present disclosure to be.

Referring to FIG. 4B, an algorithm of a power model (optimal power model B) for outputting an expected power consumption of a display unit corresponding to the sub-pixel value of an image is shown. Optimal power model B improves calculation efficiency by applying the concepts of instruction level parallelism (ILP) and thread level parallelism (TLP), and applying 1/4 or 1/8 image (image) downsampling, reducing the order of the equation, etc. It is a model with lower computational complexity.

The controller 170 receives an image as an input and performs 1/4 or 1/8 image downsampling to reduce the size of the image. The control unit 170 creates four threads, and the terms of the equation corresponding to 1 channel, 2 channels, and 3 channels for each thread (e.g., the 1 channel equation is f(R _i ), h( G _i ), the _{sum of k(B i} ), the 2-channel formula means D ₂ (R _i ,G _i ,B _i ), the 3-channel formula means D ₃ (R _i ,G _i ,B _i ), The two-channel equation is based on Equations 4 to 7, and the three-channel equation is based on Equations 8 to 9). At this time, it is possible to separate terms from equations corresponding to 1 channel, 2 channels, and 3 channels and distribute them to each thread. For example, there may be a total of 9 terms for each channel term, and it may be calculated by taking 2 to 3 terms for each thread. As each term is computed, it is performed iteratively on the pixels, which is done by a loop statement. In connection with execution of the loop statement, the controller 170 may perform instruction-level parallelization by applying NEON instructions. (Refer to Fig. 5 for instruction-level parallelization.) When the calculations in all terms are completed, the four threads are terminated, and the display power is calculated using the previously calculated values (1ch_current, 2ch_current, 3rd_current). I can.

FIG. 5 is a diagram for describing a method for analyzing power consumption of a display unit based on parallelization of a command level, according to an embodiment of the inventive concept.

Referring to FIG. 5, as an example, the controller 170 may perform performance optimization through command level parallelism. At this time, NEON instructions can be used. NEON is a 64/128 bit SIMD instruction set that can be used in a mobile application processor (AP) with an ARM core. NEON instructions can minimize the number of instructions required for calculation by implementing the SIMD architecture inside the mobile device. For example, the mobile device may be a Galaxy S3, which may be a device equipped with an ARM Cortex-A9 architecture, and thus can apply NEON.

As described above, in the power model, there are many iterative tasks performed for each pixel, and since this is performed in a loop (by a loop statement), the control unit 170 applies the NEON command to all loops existing in the power model. You can optimize the time to perform the task.

As an example, referring to FIG. 5, the NEON instruction can be applied in a loop in which the loop unrolling degree is set to 4, and with only one repetition, four values (β(R _i ,B _i ) ~ β(R _i+3 , Since B _i+3 )) can be obtained, calculations can be performed efficiently.

As a result, when the instruction level parallelization using NEON is applied to all terms of the power model, the task execution time can be greatly shortened.

FIG. 6 is a diagram illustrating a method for analyzing power consumption of a display unit based on parallelization of a thread level, according to an embodiment according to the technical idea of the present disclosure.

The control unit 170 may increase the calculation efficiency of equations required for the power model through thread-level parallelization. For example, the control unit 170 may increase the calculation efficiency of terms in the equation required for the power model through thread-level parallelization using POSIX threads. Here, since the POSIX thread is a standard API provided for implementing parallel processing software in Unix-based operating systems, multiple threads can be implemented at the operating system level inside the Android mobile device using POSIX threads.

For example, the mobile device may be a Galaxy S3, which is equipped with the ARM Cortex-A9 architecture, and thus, the control unit 170 can configure up to 4 threads, and if all threads are used, the calculation is up to 72.7%. You can increase the efficiency.

6, the control unit 170 indicates that the terms in the equation (f(R _i ) ~ x ₅ (B _i )R _i 5) (z(Ri,Bi) of the 3-channel equation is ^{5th order in four threads.} Assumptions) are appropriately distributed in parallel, and calculations can be performed in a short time by calculating based on this.

7A is an analysis of power consumption of the display unit by limiting the polynomial order of the 2-channel dependent current compensation term and the 3-channel dependent current compensation term of the power model in order to minimize computational complexity according to an embodiment of the inventive concept of the present disclosure. It is a figure for demonstrating the method for doing.

The controller 170 may perform calculation by lowering the order of an equation used in the power model in order to minimize the high computational complexity of the power model. If the order of the formula is determined to be the maximum order of 5, the calculation execution time is greatly increased in the two-channel and three-channel formulas with high calculation complexity. As shown in Equations 4 to 9, since the calculation is performed for each pixel in the image, the terms of the high order require a large amount of calculation for each iteration, and thus, a large amount of total work execution time is owned. Therefore, if the order of the terms in the formulas of 2 and 3 channels (2-channel dependent current compensation term and 3-channel dependent current compensation term) is reduced from 5 to 3, the amount of calculation required for each iteration is reduced, and the total calculation amount is up to 51%. Can be reduced.

7B is an analysis of power consumption of the display unit by limiting the polynomial order of the two-channel dependent current compensation term and the three-channel dependent current compensation term of the power model in order to minimize computational complexity according to an embodiment of the inventive concept of the present disclosure. It is a diagram for explaining the effect of the following method.

Referring to FIG. 7B, an error rate according to the order of the power model equation is shown for 256 monochrome images and 133 general images in four image databases (Caltech-256, CSIQ, TOYAMA, LIVE). In general, when using a high-order formula, the accuracy of the model increases, but in reality, the formula with the fifth order is overfitting for monochromatic images, resulting in a low error rate, but the error rates for general images are three-order and fourth-order formulas. Higher than That is, even if the order of the formula is set to be high, the error rate for the general image is not low, so the control unit 170 may set the order of the two-channel formula and the three-channel formula to 3. Referring to FIG. 7A, the control unit 170 _{lowers the order of α(R i} ,G _i ), β(R _i ,G _i ), γ(R _i ,G _i ) from 5 to 3, and z _k (Ri ,Bi) and x _u (B _i )R _i ^u _{The coefficient of each term (G i} ³ _, G _i ² , G _i , R _i ³ _, R _i ² , R _i terms) can be determined by lowering the degree of from 5 to 3. However, in the case of a 1-channel equation, the control unit 170 may be pre-calculated and store the result in a lookup table, and it may not be necessary to optimize the equation by lowering the order of the equation.

In addition, the controller 170 may perform calculation by down-sampling the image in order to minimize the high computational complexity of the power model. For example, the image may have a size of an HD (720x1280)-class image, and when the control unit 170 calculates power consumption based on a power model using the image, power calculation is performed for all pixels in the image. Therefore, the computational complexity can be increased. Accordingly, it is possible to reduce the number of pixels required for calculating power consumption based on the power model through downsampling of the image. For example, since adjacent pixels have similar values, the control unit 170 may perform 1/4 or 1/8 downsampling by grouping (creating a pixel group) 4 or 8 pixels close to all pixels of the image. have. For example, the controller 170 may obtain a representative pixel value of each pixel group through averaging and simple selection in each pixel group. Here, the averaging is a method of using the average value of all pixels in the pixel group as a representative value, and the simple selection is a method of simply selecting one representative pixel from the pixel group. When the controller 170 applies 1/4 or 1/8 downsampling to the image, the total calculation amount of power consumption of the display unit based on the power model can be reduced by a maximum of 63.9% and 76.4%, respectively. In addition, the control unit 170 may set the loop unrolling degree to 4 or 8 in a loop performed during the downsampling process, thereby increasing the efficiency of an ARM instruction required for loop execution.

FIG. 8 is a diagram for explaining a method for analyzing power consumption of a display unit by using a lookup table with divided images according to an embodiment of the inventive concept.

The control unit 170 may reduce the amount of calculation as much as possible by using the lookup table. Referring to FIG. 8, after dividing an image into NxM regions, the controller 170 may obtain an RGB value of each divided region. At this time, N and M may be 15, 15, or 5,4, and in this case, the image may be divided into 225 or 20 regions.

In this case, the RGB value of each divided area may be an average value of RGB pixels included in each divided area. The controller 170 may obtain an index value through a process of hashing the RGB values of each divided area.

The controller 170 may access a look-up table pre-configured with the estimated power consumption value using an index to obtain a power value of each divided region. The controller 170 may calculate the finally estimated power consumption of the display unit based on the power value of each divided area.

However, in this case, the size of the lookup table to be stored in the memory may be determined in consideration of the memory usage aspect in that the lookup table is a 3D array set based on three subpixel values. Since three sub-pixel values range from 0 to 255, a lookup table that stores power consumption for each sub-pixel value uses an excessive amount of memory of (256x256x256) x 8 bytes, that is, 16,777,216 x 8 bytes. Accordingly, in order to reduce the memory usage of the lookup table, the size of the lookup table can be reduced by dividing the range of the subpixel values 0 to 255 by an integer h greater than 1. For example, when h is set to 30, 10, or 5, the memory usage may be reduced by 99% or more. In this case, the lookup table determined according to the value of h may be referred to as LUTh.

In combination with the above optimization methods, the calculation complexity minimization methods, the control unit 170 can reduce the execution time of calculation using the power model to the maximum.

For example, the control unit 170 may generate a lookup table using a power consumption value estimated by a power model in which the order of terms is lowered. The controller 170 may reduce the size of the image by performing 1/4 or 1/8 downsampling on the image. The controller 170 may divide the reduced image into NxM regions, hash the RGB values of each divided region, and convert the reduced image into an index value of the lookup table. The controller 170 may access the lookup table with the index of each divided region to obtain a corresponding power value, and calculate (estimate) power consumption of the display unit based on the power value of each divided region. In this case, the mobile device can not only estimate the power consumption of the display unit with high accuracy, but also calculate the power consumption of the display unit in real time.

9 is a diagram illustrating a software platform for implementing a method of analyzing power consumption of a display unit of an electronic device according to an embodiment of the inventive concept.

The software platform shown in FIG. 9 is an example of a form in which each component of the control unit 170 is implemented in the electronic device 100 as a function unit, assuming that the electronic device 100 is operated by the Android operating system. Please let me know that it will be.

Referring to FIG. 9, a software platform for implementing a method of analyzing power consumption of a display unit of an electronic device according to an embodiment of the inventive concept is illustrated. For example, the mobile device Galaxy S3 is equipped with a Cortex-A9 CPU with an Exynos 4412 application processor with a built-in quad core. In order to port the power model (software for implementing the power consumption analysis method) to mobile devices, CyanogenMod 13.0, an open source firmware distribution for smartphones and tablets based on the Android mobile operating system, is installed, and the Android version is Marshmallow 6.0.1. Can be used.

Referring to FIG. 9, the controller 170 may transmit an image to SurfaceFlinger to generate a screen, and prepare to display an image on the screen. The controller 170 may transmit the currently displayed image to a buffer using the HWComposer (a core program of the Android system that synthesizes multiple screens and exports them to a frame buffer) inside the SurfaceFlinger. That is, while the currently displayed image is being transmitted from the front buffer to the frame buffer, the control unit 170 generates data using SurfaceFlinger and displays the next image in the back buffer. The control unit 170 flips the front and back buffers using the fb_post() function of the Hardware Abstraction Layer (HAL) and transfers the data to the front buffer and writes the data to the frame buffer. The controller 170 may finally estimate a display power value by performing an optimal power model calculation using image pixel values in the frame buffer.

The control unit 170 may implement an optimal power model inside the fb_post() function, execute in real time based on the implemented function, and automatically execute whenever a frame is changed.

10 is a table comparing effects of a method for analyzing power consumption of a display unit of an electronic device based on a power model according to various embodiments of the present disclosure.

Referring to FIG. 10, power consumption calculation times and error rates of the Hong model and the optimal power models A (see FIG. 4A) and B (see FIG. 4B) in the Galaxy S3 panel are shown.

The power consumption calculation time based on the power model is the time measured for 133 images in four image databases (Caltech-256, CSIQ, TOYAMA, LIVE). The average power consumption calculation time of the optimal power models A and B may be 0.0077 seconds and 0.11 seconds. This is 384.4 times and 26.9 times faster than the power consumption calculation time of the Hong model, respectively. That is, when the control unit 170 calculates the power consumption using the optimal power model, the power consumption calculation time may be significantly shortened compared to the case of using the Hong model. In particular, since the optimal power model A uses a look-up table that is pre-configured with the power calculation values of the power model, there is no load generated during power calculation, and thus the power consumption calculation time can be significantly shortened.

Optimal power model B exhibits an error rate similar to that of the Hong model, but optimal power model A exhibits an error rate that is about 0.7% larger than that of the Hong model. In the case of the optimal power model A, the image is divided into NxM regions, and a lookup table implemented in advance is used. Since the sampling interval between subpixels is determined according to the values of N and M that determine the image segmentation size, and the value of h determines the size of the lookup table, the error rate varies according to the values of N, M, and h. The smaller N and M and the larger h, the larger the error rate, and vice versa, the smaller the error rate.

On the other hand, the error rate of other models (for example, Park et al., Accurate power model for mobile AMOLED displays) for 133 images in 4 databases on the Galaxy S3 panel is significantly higher than that of the optimal power model A. Therefore, the optimal power model A shows excellent performance in terms of power consumption calculation time and accuracy.

When a mobile device uses an application program that changes screens quickly, such as playing games and dynamic images, the refresh rate of the screen is 60Hz. In order to apply the power model to the environment of using such an application program, the power consumption must be calculated within 0.01 seconds (10 milli-seconds) inside the system of the mobile device.

Therefore, the optimal power model A may be the most suitable power model for real-time use in a mobile system because it has an excellent error rate of 1.63% and an average performance of 0.0077 seconds (7.7 milli-seconds).

The optimal power model B has an average power consumption calculation time of 0.11 seconds and may need to be supplemented to be used in real time, but may be a power model that can be sufficiently effectively utilized in quasi-real-time analysis and processing due to a very low error rate and fast calculation speed.

11 is a graph comparing average execution times of a method for analyzing power consumption of a display unit of an electronic device based on various settings of a power model according to an exemplary embodiment of the inventive concept.

Referring to FIG. 11, a graph in which the electronic device 100 analyzes power consumption of the display unit according to various settings of the optimal power model A and compares the analysis time of the power consumption. That is, FIG. 11 shows the analysis time of the average power consumption of the optimal power model A for four databases (Caltech-256, CSIQ, TOYAMA, LIVE) when the image downsampling and NxM segmentation settings are changed.

In this case, since the lookup table changed according to the value of h is not related to the analysis time of the average power consumption, it is assumed that the lookup table 10 (LUT10) in which h is set to 10 is used. N×M is set to 5×4 or 15×15, and the change in execution time according to each setting is not large.

When image downsampling is not performed, the size of the input image (720x1280) is maintained as it is, and the computational amount increases accordingly, and thus the slowest power consumption analysis time is shown. The fastest case is the case of applying 1/8 image downsampling (simple selecting) with a loop unrolling degree of 8.

The average power consumption calculation time is 0.0078 seconds in the case of 5x4 divisions, and 0.0077 seconds in the case of 15x15 divisions.

Referring to FIG. 11, it can be seen that 1/8 image downsampling shows a faster execution time because it can reduce a computational amount compared to 1/2 image downsampling or 1/4 image downsampling. In addition, when downsampling, the averaging method may show a slower execution time than the simple selecting method. This is because a considerable amount of time occurs in the process of averaging very many pixels. Finally, if you set loop unrolling for the averaging method or simple selection method in the downsampling process to increase the efficiency of the command, you can shorten the time by up to two times.

12 is a graph comparing an average error rate of a method of analyzing power consumption of a display unit of an electronic device based on various settings of a power model according to an exemplary embodiment of the inventive concept.

12 shows error rates for lookup table 30 (LUT30), lookup table 10 (LUT10), and lookup table 5 (LUT5) implemented differently according to h. In particular, FIG. 12 shows error rates based on the various lookup tables in four cases implemented by a combination of segmentation of N×M regions of an image and 1/8 image downsampling.

At this time, LUT10 shows the lowest error rate and LUT30 shows the highest error rate. In the case of LUT30, since h, which is the sampling interval of sub-pixel values, is set to 30, the error rate is worse than that of LUT10 and LUT5, which are divided more finely. In the case of LUT10, since the sampling interval of sub-pixel values is 10, the error rate of the analyzed power is not large and shows the best result.

In the case of LUT5, the sampling interval of sub-pixel values is the smallest, but since the look-up table is implemented with the power estimated by the power model, the error rate is slightly higher than that of LUT10 due to the effect of the error rate of the model.

In addition, when the input image is divided into 15×15 areas, the error rate is lower than when the input image is divided into 5×5 areas, and whether or not image downsampling is applied has a slight effect on the error rate.

Therefore, according to the optimal power model A, it can be confirmed that the calculation time of power consumption is quite fast and high accuracy is maintained (low error rate), and thus, it is a model suitable for real-time application in an actual mobile system.

As a result, the method for analyzing power consumption of a display panel according to various embodiments of the technical idea of the present disclosure improves the limitation that it is difficult to execute a real-time power model because a mobile device such as a smartphone has lower CPU and GPU performance than a general-purpose computer. It is of great significance in that it can be executed in real time on the Android smartphone platform.

Further, according to the method for analyzing power consumption of a display panel according to various embodiments of the inventive concept, a dynamic voltage variable technique, which is a method for reducing the power consumption occupied by a mobile AMOLED display in an embedded system with limited battery usage, is performed or It can be widely used, such as being able to predict the remaining battery capacity by using it in the calculation of a battery model, and can greatly contribute to the development of low power technology.

So far, the technical idea of the present disclosure has been looked at around preferred embodiments. Those of ordinary skill in the art to which the present disclosure pertains will appreciate that the present disclosure may be implemented in a modified form without departing from the essential characteristics of the present disclosure. Therefore, the published embodiments should be considered from a descriptive point of view rather than a limiting point of view. The scope of the technical idea of the present disclosure is indicated in the claims rather than the above description, and all differences within the scope equivalent thereto should be interpreted as being included in the technical idea of the present disclosure.

Furthermore, although each drawing has been described separately for convenience of description, it is also possible to design a new embodiment by merging the embodiments described in each drawing.

In addition, the above-described technical idea of the present disclosure can be implemented as a computer-readable code on a medium in which a program is recorded. The computer-readable medium includes all types of recording devices that store data that can be read by a computer system. Examples of computer-readable media include hard disk drives (HDDs), solid state disks (SSDs), silicon disk drives (SDDs), ROMs, RAM, CD-ROMs, magnetic tapes, floppy disks, optical data storage devices, etc. Includes.

100: electronic device 120: output
130: user input unit 140: power supply unit
150: memory 160: interface unit
170: control unit

Claims (14)

As a method of analyzing the power consumption of the display panel,
Downsampling at least one image included in image data displayed on the display panel;
Generating a power consumption model of the display panel in consideration of a current dependence relationship between sub-pixels included in the display panel; And
Calculating power consumption of the display panel based on the downsampled at least one image using the power consumption model of the display panel;
Including, the method.
The method of claim 1,
The power consumption model of the display panel is expressed as a product of a current of the display panel and a voltage of the display panel,
The current of the display panel is based on a current of each of the first to third sub-pixels-hereinafter referred to as'channel independent current'-according to a dependency relationship between two sub-pixels of the first to third sub-pixels. Current-hereinafter referred to as'two-channel current dependence'-is subtracted, and current according to the dependence relationship between the first to third sub-pixels-hereinafter referred to as'three-channel current dependence'-is added, and the display The method, expressed as adding a fixed current of the panel.
The method of claim 2,
The two-channel current dependence relationship is expressed by a first polynomial for any one of the two sub-pixels,
The method of claim 1, wherein the first polynomial is a polynomial of less than 5th order.
The method of claim 3,
The method of claim 1, wherein the maximum order of the first polynomial is of third order.
The method of claim 2,
The three-channel current dependence relationship is expressed by a second polynomial for any one of the first to third sub-pixels,
Wherein the second polynomial is a polynomial of less than 5th order.
The method of claim 5,
The method of claim 1, wherein the second polynomial has a maximum order of third order.
The method of claim 2,
Calculating the power consumption of the display panel,
Calculating any one of the channel independent current, the two-channel current dependence and the three-channel current dependence for at least two or more of the plurality of pixels of the downsampled at least one image in parallel .
The method of claim 2,
Calculating the power consumption of the display panel,
And calculating in parallel by dividing at least two of the channel independent current, the two-channel current dependence and the three-channel current dependence for each of a plurality of pixels of the downsampled at least one image.
The method of claim 1,
Downsampling the at least one image,
Grouping and downsampling adjacent pixels of the at least one image.
The method of claim 9,
Grouping and downsampling adjacent pixels of the at least one image,
A method of calculating an average value of adjacent pixels in each group or selecting and downsampling a representative pixel from among adjacent pixels in each group.
The method of claim 1,
Calculating the power consumption of the display panel,
Dividing the downsampled at least one image into a predetermined number of regions;
Determining index values of a lookup table based on an average value of subpixels of each of the divided regions;
Accessing the lookup table based on index values of the lookup table to obtain power consumption for the partitioned regions; And
Calculating power consumption of the display panel based on power consumption of the divided regions;
Including, the method.
The method of claim 11,
The lookup table stores a value obtained by pre-calculating the power consumption of the divided regions based on the power consumption model of the display panel.
The method of claim 11,
Determining index values of the lookup table,
Hashing by dividing an average value of sub-pixels of each of the divided regions by a predetermined integer; And
Determining index values of the lookup table based on the hashing result;
Including, the method.
Display panel; And
Includes; a control unit for controlling the display panel,
The control unit,
A downsampling unit performing downsampling on at least one image included in image data displayed on the display panel;
A power consumption model generator configured to generate a power consumption model of the display panel in consideration of a current dependence relationship between subpixels included in the display panel; And
And a power consumption calculator configured to calculate power consumption of the display panel based on the down-sampled at least one image using the power consumption model of the display panel.

Images