Apparatus and method for evaluating display panel degradation and display driver using degradation evaluation value

Abstract

Disclosed are an apparatus and method for evaluating a deterioration state of a display panel, such as unevenness. The method for evaluating the deterioration of the display panel can be implemented by generating mutual information by using a first histogram distribution vector of a reference frame having a target gray level and a second histogram distribution vector of an evaluation frame displayed on the display panel in response to the target gray level, generating normalized mutual information of the mutual information, providing a weight incorporating perceptual features of a gray level distribution of pixels of the evaluation frame, and outputting an evaluation value obtained by multiplying the normalized mutual information by the weight.

Description

Technical Field

[0001] This disclosure relates to the assessment of degradation of a display panel, and more specifically, to apparatus and method for assessing degradation of a display panel, for assessing degradation states of a display panel such as mura, and a display driver using the degradation assessment value. Background Technology

[0002] An ideal display panel responds to display data with a target gray level by outputting an image with the same gray level as the target gray level.

[0003] However, when degradation such as unevenness occurs in the display panel, an image with a gray level different from the target gray level is output through the degraded pixels or areas of the display panel.

[0004] In most cases, the degree of degradation, such as unevenness, can be determined by having workers directly visually assess the display panel.

[0005] When degradation exists in the display panel, the display system employs compensation techniques to address degradation such as unevenness. Therefore, the display panel outputs an image with reduced degradation in response to display data corrected through compensation techniques.

[0006] To more accurately compensate for degradation, a quantitative assessment of the display panel's degradation status is required. Furthermore, even if the degree of degradation has been reduced through the application of compensation techniques, a quantitative assessment of the display panel's degradation status, even with the reduced degradation level, is still necessary.

[0007] Therefore, it is necessary to develop technologies that can quantitatively assess the degradation status of display panels or degradation status with reduced degradation levels. Summary of the Invention

[0008] Various embodiments relate to providing apparatus and methods for assessing the degradation of a display panel, such as the degree of uneven degradation, which can quantitatively evaluate the degradation of the display panel.

[0009] Furthermore, various embodiments relate to providing an apparatus and method for assessing the degradation of a display panel by analyzing the correlation between a reference frame of a reference image and an evaluation frame of an evaluation image.

[0010] Furthermore, various embodiments relate to providing a display driver capable of compensating for display panel degradation by using evaluation values ​​obtained through assessing the degradation state to resolve display panel degradation.

[0011] In one embodiment, the apparatus for evaluating the degradation of a display panel includes: a first histogram analysis unit configured to output a first histogram distribution vector of a reference frame having a target gray level; a second histogram analysis unit configured to output a second histogram distribution vector of an evaluation frame displayed on the display panel in response to the target gray level; a correlation analysis unit configured to generate mutual information by using the first and second histogram distribution vectors, and to generate normalized mutual information of the mutual information; a weight providing unit configured to provide weights of cognitive features of the gray level distribution of pixels incorporating the evaluation frame; and an output unit configured to output an evaluation value obtained by calculating the normalized mutual information and the weights.

[0012] In one implementation, the method for evaluating the degradation of a display panel includes: outputting a first histogram distribution vector of a reference frame having a target gray level; outputting a second histogram distribution vector of an evaluation frame displayed on the display panel in response to the target gray level; generating mutual information using the first and second histogram distribution vectors; generating normalized mutual information of the mutual information; providing weights of cognitive features of the gray level distribution of pixels incorporating the evaluation frame; and outputting an evaluation value obtained by multiplying the normalized mutual information by the weights.

[0013] In one implementation, the display driver includes: an evaluation value storage unit configured to store evaluation values; and a degradation compensator configured to receive the evaluation values ​​and compensate for degradation by converting display data based on the evaluation values. The evaluation values ​​correspond to values ​​obtained by calculating normalized mutual information and weights.

[0014] The advantage of this disclosure is that it can calculate quantitative evaluation values ​​in response to conditions such as uneven degradation of the display panel.

[0015] Furthermore, the advantage of this disclosure is that it can determine the degree of degradation of the display panel based on a quantitative evaluation value calculated by analyzing the correlation between the reference frame of the reference image and the evaluation frame of the evaluation image.

[0016] Furthermore, the advantage of this disclosure is that it can address the degradation of the display panel by using evaluation values ​​obtained through assessing the degradation status to compensate for the display data. Attached Figure Description

[0017] Figure 1 This is a block diagram illustrating an apparatus for evaluating the degradation of a display panel according to a preferred embodiment of the present disclosure.

[0018] Figure 2 This is an example of a histogram of the reference frame.

[0019] Figure 3 This is an example of a histogram for evaluating frames.

[0020] Figure 4 This is an example diagram illustrating the relationship between entropy and mutual information in the form of a frequency band diagram.

[0021] Figure 5 This is an example image of local pixels surrounding a pixel.

[0022] Figure 6 It is a graph showing the relationship between just-perceptible differences and average brightness values.

[0023] Figure 7 This is a block diagram showing the display system.

[0024] Figure 8 This is a detailed block diagram of the display driver disclosed herein. Detailed Implementation

[0025] This disclosure discloses a technique for outputting quantitative results for evaluating the degree of degradation, such as unevenness, present in a display panel (not shown).

[0026] This disclosure is configured to output an evaluation value to assess the degree of degradation by analyzing the correlation between a reference frame of a reference image and an evaluation frame of an evaluation image.

[0027] In this context, the reference image can be understood as a pre-set image used for comparison with the evaluation image. A frame of the reference image is called the reference frame. All pixels in the reference frame are set to have the same target gray level.

[0028] An evaluation image can be understood as an image displayed on a display panel. A frame of the evaluation image is called an evaluation frame. An evaluation frame can be understood as an image whose display data, with a target grayscale level, is displayed on the display panel.

[0029] In this case, the target gray level of the reference frame and the target gray level that provides the display data for displaying the evaluation frame can be understood as being the same.

[0030] This disclosure is configured to assess the degradation of the evaluation image by analyzing the correlation between the reference frame and the evaluation frame, and output the evaluation value as the result of the evaluation.

[0031] The embodiments of this disclosure can be as follows: Figure 1 The structure shown. Figure 1 The implementation includes a first bar chart analysis unit 10, a second bar chart analysis unit 20, a correlation analysis unit 30, a weight provision unit 40, and an output unit 50.

[0032] The first histogram analysis unit 10 outputs the first histogram distribution vector X of the reference frame with the target gray level.

[0033] To this end, the first histogram analysis unit 10 generates a graph such as the number of pixels per gray level in the reference frame. Figure 2 The first histogram shows that the gray levels of the reference frame pixels are distributed only within the target gray levels.

[0034] The first histogram analysis unit 10 can output a histogram distribution vector X representing the changes in pixel distribution for each gray level, such as... Figure 2 .

[0035] The second histogram analysis unit 20 outputs the second histogram distribution vector Y of the evaluation frame displayed on the display panel in response to the target gray level.

[0036] To this end, the second histogram analysis unit 20 generates a representation such as the number of pixels per gray level in the evaluation frame. Figure 3 The second histogram shows that the gray levels of the pixels in the evaluation frame are distributed around the target gray level. Pixels in the evaluation frame with the target gray level can be considered as having no degradation. Pixels outside the target gray level in the evaluation frame can be considered as having gray levels distorted due to degradation.

[0037] The second histogram analysis unit 20 can output a histogram distribution vector Y representing the changes in pixel distribution for each gray level, such as... Figure 3 .

[0038] The correlation analysis unit 30 is configured to generate mutual information (hereinafter referred to as "MI") by using the first histogram distribution vector X and the second histogram distribution vector Y, and to generate normalized mutual information (hereinafter referred to as "NMI") of the mutual information MI.

[0039] For this purpose, the correlation analysis unit 30 may include an MI generator 32 and an NMI generator 34.

[0040] In the mutual information MI generator 32 and the NMI generator 34, the mutual information MI generator 32 can generate mutual information MI according to the following equation 1.

[0041] [Equation 1]

[0042]

[0043] In Equation 1, MI(X, Y) represents the mutual information between the first histogram distribution vector X and the second histogram distribution vector Y. x is the discrete probability variable of the first histogram distribution vector X. y is the discrete probability variable of the second histogram distribution vector Y. p(x, y) is the joint probability distribution of the discrete probability variables x and y. p(x) is the surrounding probability distribution of the discrete probability variable x of the histogram distribution vector X of the reference frame. p(y) is the surrounding probability distribution of the discrete probability variable y of the histogram distribution vector Y of the reference frame.

[0044] Mutual information MI(X,Y) represents the amount of information about the relationship between discrete probability variables x and y. More specifically, mutual information MI(X,Y) refers to the interdependence between discrete probability variables x and y, that is, the amount of information provided by one probability variable relative to another.

[0045] When the joint probability distribution of discrete probability variables x and y is p(x, y) and the surrounding probability distributions of discrete probability variables x and y are p(x) and p(y), the mutual information MI(X, Y) can be calculated according to Equation 1.

[0046] Mutual information MI(X, Y) can be generated using entropy according to Equation 2 below. In this case, entropy refers to the expected value of the information content of all cases and is used to represent the magnitude of the uncertainty of the probability distribution of all cases.

[0047] [Equation 2]

[0048] MI(X,Y)=H(X)+H(Y)-H(X,Y)

[0049] In Equation 2, MI(X, Y) represents the mutual information between the first histogram distribution vector X and the second histogram distribution vector Y. H(X) is the entropy of the first histogram distribution vector X and represents the gray level distribution of the reference frame. H(Y) is the entropy of the second histogram distribution vector Y and represents the gray level distribution of the evaluation frame. H(X, Y) is the joint entropy and represents the sum of the gray level distributions of the reference frame that do not overlap with the gray levels of the evaluation frame and the gray level distributions of the evaluation frame that do not overlap with the gray levels of the reference frame.

[0050] In Equation 2, H(X) is the Shannon entropy of p(x). H(Y) is the Shannon entropy of p(y). The Shannon entropy H(X) can be calculated using Equation 3 below. The Shannon entropy H(Y) can be calculated using Equation 4 below.

[0051] [Equation 3]

[0052] H(X) = -Σ x p(X)*logp(X)

[0053] [Equation 4]

[0054] H(Y)=-∑ y p(Y)*logp(Y)

[0055] In equations 3 and 4, p(X) is the probability distribution of the first histogram distribution vector X. p(Y) is the probability distribution of the second histogram distribution vector Y. x is the discrete probability variable of the first histogram distribution vector X. y is the discrete probability variable of the second histogram distribution vector Y.

[0056] Equations 2, 3, and 4 can be found in Bouma, Gerlof's "Normalized (pointwise) Mutual Information in Collocation Extraction," which is presented in the Proceedings of GSCL 31-40 in 2009.

[0057] The relationship between entropy and mutual information MI(X,Y) can be shown as follows: Figure 4 The frequency band diagram.

[0058] exist Figure 4 In this context, the mutual information MI(X,Y) corresponds to the intersection of the Shannon entropy H(X) and the Shannon entropy H(Y). Furthermore, the joint entropy H(X,Y) corresponds to the union of H(X|Y) and H(Y|X), where H(X|Y) does not overlap with the evaluation frame in the Shannon entropy H(X) of the reference frame, and H(Y|X) does not overlap with the reference frame in the Shannon entropy H(Y) of the evaluation frame.

[0059] according to Figure 4 To understand this more clearly, mutual information MI(X,Y) is generated using the entropy according to Equation 2.

[0060] The potential problem with mutual information MI(X,Y) is that it is used for clustering and classification, and therefore requires general normalization.

[0061] NMI generator 34 is used to normalize mutual information MI(X, Y) and is configured to convert mutual information MI(X, Y) into NMI with a preset range. In this case, the preset range can be, for example, 0 to 1, for normalizing NMI.

[0062] NMI generator 34 can convert MI to NMI according to the following equation 5.

[0063] [Equation 5]

[0064]

[0065] In Equation 5, Shannon entropy H(X) is calculated using Equation 3 above, and Shannon entropy H(Y) is calculated using Equation 4 above.

[0066] NMI can be expressed as a value between 0 and 1 according to Equation 5 above. When NMI is close to 1, the similarity between the first histogram distribution vector X and the second histogram distribution vector Y is high. When NMI is close to 0, the difference between the first histogram distribution vector X and the second histogram distribution vector Y is large.

[0067] Output unit 50 modifies the NMI provided by NMI generator 34 of correlation analysis unit 30, thereby incorporating the cognitive characteristics of the gray-level distribution of pixels in the evaluation frame into the NMI, and outputs an evaluation value corresponding to the modification. Cognitive characteristics can be understood as quantifying the degree to which a human perceives the gray-level distribution of pixels in the evaluation frame.

[0068] Therefore, the output unit 50 is configured to output the evaluation value obtained by multiplying the NMI by the weight WT of the weight providing unit 40.

[0069] The weights WT of the cognitive features of the gray-level distribution of pixels used to be incorporated into the evaluation frame are provided by the weight providing unit 40.

[0070] To this end, the weight providing unit 40 includes a just-perceptible difference (hereinafter referred to as "JND") generator 42, a JND average value generator 44, and a weight generator 46.

[0071] In this case, JND defines the minimum perceived brightness difference of the average brightness value based on the average brightness value among the local features of the image signal.

[0072] The JND is generated by the JND generator 42. That is, the JND generator 42 generates the JND of the evaluation frame for each pixel.

[0073] More specifically, the JND generator 42 first calculates the average brightness value of the local pixels surrounding each pixel for each pixel in the evaluation frame.

[0074] The average brightness value of local pixels around coordinates (x, y) can be defined as I(x, y). Figure 5 In this context, the average brightness value I(x, y) can be understood as the average brightness value of the pixels belonging to the five columns and five rows surrounding the coordinate (x, y).

[0075] JND generator 42 calculates the average local pixel brightness value of all pixels in the evaluation frame.

[0076] In addition, JND generator 42 provides JND corresponding to the average brightness value.

[0077] The coordinates (x, y) can be recognized by a human when the brightness of the pixel at coordinate (x, y) differs from I(x, y) (i.e., the average value of the local pixels) by JND or more. Conversely, the coordinates (x, y) cannot be recognized by a human when the brightness of the pixel at coordinate (x, y) differs from I(x, y) (i.e., the average value of the local pixels) by less than JND.

[0078] That is, JND can be understood as the threshold at which it can be perceived by humans, and can be like... Figure 6 The settings shown are for the average brightness value I(x, y).

[0079] Reference Figure 6 Based on the middle of the average brightness value range, JND has a value that increases along the first curve as the average brightness value increases, and a value that increases along the second curve as the average brightness value decreases. In this case, the second curve has a higher growth rate than the first curve.

[0080] For example, within the range of average brightness values ​​defined between 0 and 255, the sensitivity is lowest in the central brightness region surrounding 128. Furthermore, sensitivity increases along the first curve as the brightness gradually increases from the average brightness value of 128. Sensitivity increases along the second curve as the brightness gradually decreases from the average brightness value of 128. Moreover, the change in sensitivity in the region where the brightness gradually decreases from the average brightness value of 128 is greater than the change in sensitivity in the region where the brightness gradually increases from the average brightness value of 128.

[0081] JND generator 42 can provide, for example, Figure 6 The JND value corresponding to the average brightness value of the pixels in the evaluation frame shown.

[0082] JND average value generator 44 generates the average value JNDm of the JND of the evaluation frame.

[0083] In addition, the weight generator 46 provides the weight WT corresponding to the average value JNDm.

[0084] The weight generator 46 can be configured to divide the range from which the average is formed into multiple weighted ranges, assign a preset weight to each weighted range, and output the weights corresponding to the weighted range of the average in response to the average.

[0085] For example, if the average value is formed in the range of 0 to 12, the weighted range can be divided into a first weighted range where the average value is greater than 0 and less than 4, a second weighted range where the average value is greater than 4 and less than 8, and a third weighted range where the average value is greater than 8 and less than 12.

[0086] Furthermore, the weight WT for the first weighted range can be set to 0.9. The weight WT for the second weighted range can be set to 0.95. The weight WT for the third weighted range can be set to 1.

[0087] Therefore, the weight generator 46 can select one of 9.9, 9.95 and 1 based on the average value, and can output the selected value as the weight WT.

[0088] Output unit 50 can output an evaluation value obtained by multiplying the weight WT by NMI. Therefore, the evaluation value can have a value incorporating different JND features at low and high gray levels.

[0089] refer to Figures 1 to 6 The apparatus for evaluating the degradation of a display panel according to this disclosure implements a degradation evaluation method that sequentially performs the following steps.

[0090] That is, the degradation evaluation method includes the following steps: outputting a first histogram distribution vector X of a reference frame having a target gray level; outputting a second histogram distribution vector Y of an evaluation frame displayed on a display panel in response to the target gray level; generating a MI using the first histogram distribution vector X and the second histogram distribution vector Y; generating an NMI of mutual information MI; providing a weight WT of the cognitive features of the gray level distribution of pixels incorporating the evaluation frame; and outputting an evaluation value obtained by multiplying the NMI by the WT.

[0091] This disclosure allows for the calculation of evaluation values ​​that can quantitatively assess the degradation state of a display panel, such as unevenness.

[0092] When the evaluation value calculated according to this disclosure is large, the histogram distribution vectors of the reference frame and the evaluation frame are similar, and the degradation degree of the evaluation frame can be determined to be low. When the calculated evaluation value is small, the degradation degree of the evaluation frame can be determined to be high because the difference between the histogram distribution vectors of the reference frame and the evaluation frame is large.

[0093] As described above, the advantage of this disclosure is that it can assess the degree of degradation of the display panel by analyzing the correlation between the reference frame of the reference image and the evaluation frame of the evaluation image.

[0094] like Figure 7 As shown, the display data is provided by the timing controller 100. The timing controller 100 constructs a data packet PKT of the display data and provides the data packet to the display driver 110.

[0095] The display driver 110 is configured to recover display data after receiving a data packet, generate a source signal Sout corresponding to the display data, and provide the source signal Sout to the display panel 120.

[0096] Figure 7 The display driver 110 can be configured as follows, by way of example. Figure 8 As shown.

[0097] Reference Figure 8 The display driver 110 may include a data packet receiver 200, a degradation compensator 210, a source signal generator 220, a source signal output unit 230, and an evaluation value storage unit 240.

[0098] The data packet receiver 200 receives the display data packet PKT provided by the timing controller 100 and recovers the display data from the data packet PKT.

[0099] The degradation compensator 210 receives the display data from the data packet receiver 200 and the evaluation value from the evaluation value storage unit 240, and can compensate for degradation by converting the display data based on the evaluation value.

[0100] The transformation of display data based on the evaluation value can be performed in the pixel units of the display panel. The display data can be transformed by using the operation result of the evaluation value or by using the evaluation value as a coefficient in a compensation equation.

[0101] The source signal output unit 220 drives the source signal Sout in response to the degraded and compensated display data, and provides the source signal Sout to the display panel 120.

[0102] The evaluation value storage unit 240 can be constructed using a memory such as an EEPROM, and can store and provide values ​​based on... Figures 1 to 6 The evaluation value calculated according to the embodiments of this disclosure.

[0103] As described above, the display driver of this disclosure can solve the degradation problem of the display panel by compensating the display data based on the evaluation value.

Claims

1. An apparatus for assessing the degradation of a display panel, comprising: The first histogram analysis unit is configured to output the first histogram distribution vector of the reference frame with the target gray level; The second histogram analysis unit is configured to output a second histogram distribution vector of the evaluation frame displayed on the display panel in response to the target gray level; The correlation analysis unit is configured to generate mutual information using the first histogram distribution vector and the second histogram distribution vector, and to generate normalized mutual information of the mutual information. The weighting unit is configured to provide weights that incorporate cognitive features of the gray-level distribution of pixels in the evaluation frame. as well as The output unit is configured to output the evaluation value obtained by calculating the normalized mutual information and the weights.

2. The apparatus according to claim 1, wherein, The pixels of the reference frame are configured to have the same target gray level.

3. The apparatus according to claim 1, wherein: The first histogram analysis unit generates a first histogram representing the number of pixels per gray level in the reference frame, and outputs a first histogram distribution vector corresponding to the first histogram. The second histogram analysis unit generates a second histogram representing the number of pixels per gray level in the evaluation frame, and outputs the second histogram distribution vector corresponding to the second histogram.

4. The apparatus according to claim 1, wherein, The correlation analysis unit includes: Mutual information generator, configured to utilize equations To generate the mutual information; and A normalized mutual information generator is configured to convert the mutual information into normalized mutual information with a preset range. Wherein, MI(X,Y) represents the mutual information between the first histogram distribution vector X and the second histogram distribution vector Y, x is the discrete probability variable of the first histogram distribution vector X, y is the discrete probability variable of the second histogram distribution vector Y, p(x,y) is the joint probability distribution of the discrete probability variables x and y, p(x) is the surrounding probability distribution of the discrete probability variable x, and p(y) is the surrounding probability distribution of the discrete probability variable y.

5. The apparatus according to claim 4, wherein, The normalized mutual information generator uses the equation The mutual information is converted into the normalized mutual information. Where H(X) passes through -∑ x The value of H(Y) is calculated using p(X)*logp(X), and H(Y) is calculated using -∑ y The probability distribution of the first histogram distribution vector X is calculated using p(Y)*logp(Y), where p(X) is the probability distribution of the second histogram distribution vector Y, x is the discrete probability variable of the first histogram distribution vector X, and y is the discrete probability variable of the second histogram distribution vector Y.

6. The apparatus according to claim 1, wherein, The correlation analysis unit includes: A mutual information generator is configured to generate the mutual information using the equation MI(X,Y)=H(X)+H(Y)-H(X,Y); and A normalized mutual information generator is configured to convert the mutual information into normalized mutual information with a preset range. Wherein, MI(X,Y) represents the mutual information between the first histogram distribution vector X and the second histogram distribution vector Y, H(X) is the surrounding entropy of the first histogram distribution vector X and represents the gray level distribution of the reference frame, H(Y) is the surrounding entropy of the second histogram distribution vector Y and represents the gray level distribution of the evaluation frame, and H(X,Y) is the joint entropy and represents the sum of the gray level distributions of the reference frame that do not overlap with the gray level distributions of the evaluation frame and the gray level distributions of the evaluation frame that do not overlap with the gray level distributions of the reference frame.

7. The apparatus according to claim 6, wherein, The normalized mutual information generator uses the equation The mutual information is converted into the normalized mutual information. Where H(X) passes through -∑ x The value of H(Y) is calculated using p(X)*logp(X), and H(Y) is calculated using -∑ y The probability distribution of the first histogram distribution vector X is calculated using p(Y)*logp(Y), where p(X) is the probability distribution of the second histogram distribution vector Y, x is the discrete probability variable of the first histogram distribution vector X, and y is the discrete probability variable of the second histogram distribution vector Y.

8. The apparatus according to claim 1, wherein, The weight providing unit includes: A just-perceptible difference generator is configured to generate just-perceptible differences for the evaluation frame for each pixel. A just-perceptible difference average generator is configured to generate the average of the just-perceptible differences of the evaluation frames; and A weight generator is configured to provide weights corresponding to the average value with just noticeable differences.

9. The apparatus according to claim 8, wherein: The just-perceptible difference generator calculates the average brightness value of the local pixels surrounding each pixel for each pixel, and provides the just-perceptible difference corresponding to the average brightness value. Wherein, based on the middle of the range of the average brightness values, the just-perceptible difference has a value that increases along a first curve as the average brightness value increases, and a value that increases along a second curve as the average brightness value decreases, and The second curve has a higher growth rate than the first curve.

10. The apparatus according to claim 8, wherein, The weight generator: The range from which the average value is formed is divided into multiple weighted ranges; Assign preset weights to each of the weighted ranges; and In response to the average value, the weights having the weighted range corresponding to the average value are output.

11. Methods for assessing the degradation of display panels include: Output the first histogram distribution vector of the reference frame with the target gray level; Output a second histogram distribution vector of the evaluation frame displayed on the display panel in response to the target gray level; Mutual information is generated using the first histogram distribution vector and the second histogram distribution vector; Generate normalized mutual information of the mutual information; Weights are provided for the cognitive features of the gray-level distribution of the pixels in the evaluation frame. as well as The output is an evaluation value obtained by multiplying the normalized mutual information by the weights.

12. The method according to claim 11, wherein: Outputting the first histogram distribution vector includes generating a first histogram representing the number of pixels per gray level in the reference frame, and outputting the first histogram distribution vector corresponding to the first histogram. Outputting the second histogram distribution vector includes generating a second histogram representing the number of pixels per gray level in the evaluation frame, and outputting the second histogram distribution vector corresponding to the second histogram.

13. The method according to claim 11, wherein, The mutual information is obtained by using equations To generate, Wherein, MI(X,Y) represents the mutual information between the first histogram distribution vector X and the second histogram distribution vector Y, x is the discrete probability variable of the first histogram distribution vector X, y is the discrete probability variable of the second histogram distribution vector Y, p(x,y) is the joint probability distribution of the discrete probability variables x and y, p(x) is the surrounding probability distribution of the discrete probability variable x, and p(y) is the surrounding probability distribution of the discrete probability variable y.

14. The method according to claim 11, wherein, The mutual information is generated using the equation MI(X,Y)=H(X)+H(Y)-H(X,Y). Wherein, MI(X,Y) represents the mutual information between the first histogram distribution vector X and the second histogram distribution vector Y, H(X) is the surrounding entropy of the first histogram distribution vector X and represents the gray level distribution of the reference frame, H(Y) is the surrounding entropy of the second histogram distribution vector Y and represents the gray level distribution of the evaluation frame, and H(X,Y) is the joint entropy and represents the sum of the gray level distributions of the reference frame that do not overlap with the gray level distributions of the evaluation frame and the gray level distributions of the evaluation frame that do not overlap with the gray level distributions of the reference frame.

15. The method according to claim 11, wherein, The normalized mutual information is obtained by using the equation To generate, Wherein, MI(X, Y) represents the mutual information between the first histogram distribution vector X and the second histogram distribution vector Y, and H(X) is expressed by the equation -∑ x The expression p(X)*logp(X) is used to calculate H(Y), and H(Y) is obtained through the equation -∑ y The probability distribution of the first histogram distribution vector X is calculated using p(Y)*logp(Y), where p(X) is the probability distribution of the second histogram distribution vector Y, x is the discrete probability variable of the first histogram distribution vector X, and y is the discrete probability variable of the second histogram distribution vector Y.

16. The method according to claim 11, wherein, Providing the weights includes: For each pixel, generate just-perceptible differences in the evaluation frame; Generate the average value of the just-perceptible differences of the evaluation frames; and Provide a weight corresponding to the average value of the just-perceptible difference.

17. The method according to claim 16, wherein, Generating the just-perceptible difference includes: For each pixel, calculate the average brightness value of the local pixels surrounding that pixel, and Provides a just-perceptible difference corresponding to the average brightness value. Based on the middle of the range of the average brightness values, the just-perceptible difference has a value that increases along a first curve as the average brightness value increases, and a value that increases along a second curve as the average brightness value decreases. The second curve has a higher growth rate than the first curve.

18. The method according to claim 16, wherein, Providing the weights includes: The range from which the average value is formed is divided into multiple weighted ranges; Assign preset weights to each of the weighted ranges; and In response to the average value, the weights having the weighted range corresponding to the average value are output.

19. Display driver, including: An evaluation value storage unit is configured to store evaluation values. as well as A degradation compensator is configured to receive the evaluation value and compensate for degradation by converting display data based on the evaluation value. The evaluation value corresponds to the value obtained by calculating normalized mutual information and weights. The normalized mutual information is obtained by evaluating the correlation between the histogram distribution vectors of the reference frame and the evaluation frame corresponding to the target gray level. The weights incorporate the cognitive features of the gray level distribution of the pixels in the evaluation frame.

20. The display driver according to claim 19, wherein, The normalized mutual information is generated by normalizing the mutual information, which is generated using a first histogram distribution vector of the reference frame having the target gray level and a second histogram distribution vector of the evaluation frame displayed on the display panel in response to the target gray level.

Images