Image processing method, optical compensation method, and optical compensation system

Abstract

The optical compensation method generates first image data by displaying a dot image through a display device and capturing the dot image displayed on the display device through an imaging device, generates second image data by displaying a second image through a display device and capturing the second image displayed on the display device through an imaging device, generates deblurred data by deblurring the second image data based on the first image data, and generates compensation data for brightness correction of the display device based on the deblurred data.

Description

Technology Field

[0001] The present invention relates to an image processing method, an optical compensation method, and an optical compensation system. Background Technology

[0002] As interest in information displays intensifies and the demand for portable information media rises, the demand for and commercialization of display devices are becoming a major focus. The problem to be solved

[0003] The present invention aims to provide an image processing method, an optical compensation method, and an optical compensation system capable of generating appropriate compensation data (or compensation value) for luminance correction (or stain compensation) in a display device in which the correlation between adjacent pixels is low or almost non-existent. means of solving the problem

[0004] An optical compensation method according to embodiments of the present invention comprises: a step of displaying a first image through a display device, wherein the first image includes a dot pattern in which pixels spaced apart with at least two pixels in between emit light; a step of capturing the first image displayed on the display device through an imaging device to generate first imaging data; a step of displaying a second image through the display device; a step of capturing the second image displayed on the display device through the imaging device to generate second imaging data; a step of deblurring the second imaging data based on the first imaging data to generate deblurring data; a step of generating compensation data for brightness correction of the display device based on the deblurring data; and a step of storing the compensation data in a memory device.

[0005] The first image above includes a plurality of regions divided based on a reference block, and in each of the regions, the dot pattern can be expressed by the target pixel among the pixels in the display device emitting light and the adjacent pixels among the pixels adjacent to the target pixel refraining from emitting light.

[0006] Depending on at least one of the imaging conditions and the resolution of the imaging device, the dot pattern is blurred during imaging, and the first imaging data may include a blurred dot pattern corresponding to the dot pattern.

[0007] The step of generating the deblurring data may include: a step of detecting blur information indicating the degree of blurring of the dot pattern in an image captured by the imaging device based on the first imaging data; and a step of deblurring the second imaging data based on the blur information.

[0008] The step of detecting the blur information may include the step of deriving a weight matrix for converting the blurred dot pattern included in the first image data into an ideal dot pattern.

[0009] The step of deriving the weight matrix includes the step of calculating a first weight for a target pixel corresponding to the dot pattern and a second weight for adjacent pixels adjacent to the target pixel through machine learning on the blurred dot patterns, and the first weight and the second weight may be included in the weight matrix.

[0010] The Gradient Descent Algorithm can be used in the above machine learning.

[0011] The step of deriving the weight matrix may include: a step of calculating deblurring dot data including a deblurring dot pattern using the weight matrix; a step of calculating an error between ideal dot data including an ideal dot pattern and the deblurring dot data; and a step of adjusting the first weight and the second weight based on the error.

[0012] The step of deriving the weight matrix may include the step of calculating the error and the step of adjusting the first weight and the second weight so as to minimize the error.

[0013] The step of calculating the above error can calculate the error by normalizing the first imaging data.

[0014] The step of generating the deblurring data may further include: a step of generating the first deblurring data by deblurring the second image data; a step of detecting noise through spatial frequency analysis of the first deblurring data, wherein the noise is a value deblurred outside a reference range during deblurring of the second image data; and a step of replacing the noise with a corresponding value of the second image data.

[0015] The step of generating the first image data may include the step of converting the resolution of an image captured by the image device so as to be equal to the resolution of the display device.

[0016] The compensation data stored in the memory device can be used to compensate for brightness deviation when the display device is driven.

[0017] An image processing method according to embodiments of the present invention can preprocess an image for optical compensation of a display device. The image processing method includes: a step of detecting blur information indicating the degree to which a dot pattern is blurred in a first image including a dot pattern; and a step of deblurring a second image based on the blur information.

[0018] The step of detecting the blur information may include the step of deriving a weight matrix for converting a blurred dot pattern included in the first captured image into an ideal dot pattern.

[0019] The step of deriving the weight matrix includes the step of calculating a first weight for a target pixel corresponding to the dot pattern and a second weight for adjacent pixels adjacent to the target pixel through machine learning on the blurred dot patterns, and the first weight and the second weight may be included in the weight matrix.

[0020] Gradient descent can be used in the above machine learning.

[0021] The step of deriving the weight matrix may include: generating a deblurring dot image including a deblurring dot pattern using the weight matrix; calculating an error between an ideal dot image including an ideal dot pattern and the deblurring dot image; and adjusting the first weight and the second weight based on the error.

[0022] The step of deblurring the second captured image may further include: a step of generating a first deblurred image by deblurring the second captured image; a step of detecting noise through spatial frequency analysis of the first deblurred image, wherein the noise is a value deblurred outside a reference range during deblurring of the second captured image; and a step of replacing the noise with a corresponding value of the second captured image.

[0023] An optical compensation system according to embodiments of the present invention includes: an imaging device that captures an image displayed through a display device and generates imaging data; and a brightness correction device that generates compensation data for brightness correction of the display device based on the imaging data, wherein the brightness correction device detects blur information indicating the degree to which a dot pattern is blurred in first imaging data including a dot pattern, and deblurs second imaging data based on the blur information. Effects of the invention

[0024] An image processing method, an optical compensation method, and an optical compensation system according to an embodiment of the present invention can acquire blur information indicating the degree of blur occurring during imaging based on first imaging data obtained by capturing a first image including a dot pattern, and can deblur second imaging data for a second image (i.e., an image for generating compensation data) based on the blur information. Through deblurring, the problem of blur occurring in the second imaging data can be improved and imaging accuracy can be enhanced, and more appropriate compensation data for luminance correction can be generated.

[0025] The effects according to the embodiments of the present invention are not limited to those exemplified above, and various other effects are included in this specification. Brief explanation of the drawing

[0026] FIG. 1 is a drawing showing an optical compensation system according to embodiments of the present invention. Figure 2 is a diagram illustrating the accuracy of an image captured according to the resolution of the imaging device included in the optical compensation system of Figure 1. Figure 3 is a diagram showing the luminance interference ratio between pixels according to the resolution of the imaging device. FIG. 4 is a drawing showing an embodiment of a display device included in the optical compensation system of FIG. 1. FIG. 5 is a block diagram showing one embodiment of a luminance correction device included in the optical compensation system of FIG. 1. FIG. 6 is a block diagram showing another embodiment of a luminance correction device included in the optical compensation system of FIG. 1. FIG. 7 is a drawing showing an example of a deblurring block included in the brightness correction device of FIG. 5. FIG. 8 is a drawing showing an example of first imaging data used in the deblurring block of FIG. 7. FIGS. 9 and FIGS. 10 are diagrams illustrating the operation of a weight matrix generator included in the deblurring block of FIG. 7. FIG. 11 is a diagram showing an example of a weight matrix generator included in the deblurring block of FIG. 7. Figure 12 is a diagram illustrating the process of calculating an error in the weight matrix generator of Figure 11. FIG. 13 is a diagram illustrating the operation of the arithmetic unit included in the deblurring block of FIG. 7. FIGS. 14 and FIGS. 15 are drawings showing other embodiments of a deblurring block included in the luminance correction device of FIG. 5. FIG. 16 is a diagram illustrating the operation of a noise remover included in the deblurring block of FIG. 14 and FIG. 15. FIGS. 17 to 19 are drawings illustrating the effect of deblurring on image data. FIGS. 20a and FIGS. 20b are circuit diagrams illustrating an example of a pixel included in the display device of FIG. 4. FIG. 21 is a perspective view illustrating a light-emitting element included in the pixel of FIG. 20a. FIG. 22 is a flowchart illustrating an optical compensation method according to embodiments of the present invention. FIG. 23 is a flowchart explaining the process of deblurring the second image data of FIG. 22. FIG. 24 is a flowchart explaining the process of deriving the weight matrix of FIG. 23. Specific details for implementing the invention

[0027] The present invention is capable of various modifications and may take various forms, and specific embodiments are illustrated in the drawings and described in detail in the text. However, this is not intended to limit the invention to the specific disclosed forms, and it should be understood that the invention includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the invention.

[0028] In describing each drawing, similar reference numerals have been used for similar components. In the attached drawings, the dimensions of the structures are depicted enlarged from their actual size for clarity of the invention. Terms such as "first," "second," etc., may be used to describe various components, but said components should not be limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the invention, the first component may be named the second component, and similarly, the second component may be named the first component.

[0029] In this application, terms such as "comprising" or "having" are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0030] Some embodiments are described in the attached drawings in relation to functional blocks, units, and / or modules. Those skilled in the art will understand that these blocks, units, and / or modules are physically implemented by logic circuits, individual components, microprocessors, hardwired circuits, memory elements, wiring connections, and other electronic circuits. These may be formed using semiconductor-based manufacturing techniques or other manufacturing techniques. In the case of blocks, units, and / or modules implemented by microprocessors or other similar hardware, they may be programmed and controlled using software to perform the various functions discussed in the present invention, and may optionally be driven by firmware and / or software. Additionally, each block, unit, and / or module may be implemented by dedicated hardware, or by a combination of dedicated hardware performing some functions and a processor performing other functions (e.g., one or more programmed microprocessors and associated circuits). Furthermore, in some embodiments, blocks, units, and / or modules may be physically separated into two or more individual blocks, units, and / or modules that interact within the scope of the concept of the present invention without departing from the scope of the present invention. Additionally, in some embodiments, blocks, units, and / or modules may be combined into physically more complex blocks, units, and / or modules within the scope of the concept of the present invention.

[0031] Hereinafter, preferred embodiments of the present invention and other matters necessary for those skilled in the art to easily understand the contents of the present invention will be described in detail with reference to the attached drawings. In the following description, singular expressions include plural expressions unless the context clearly indicates that only singular expressions are included.

[0032] FIG. 1 is a diagram showing an optical compensation system according to embodiments of the present invention. FIG. 2 is a diagram explaining the accuracy of an image captured according to the resolution of an imaging device included in the optical compensation system of FIG. 1. FIG. 3 is a diagram showing the luminance interference ratio between pixels according to the resolution of an imaging device.

[0033] First, referring to FIG. 1, the optical compensation system (10) (or luminance correction system, image processing system) may include an imaging device (3000) and a luminance correction device (2000) (or image processing device). According to an embodiment, the optical compensation system (10) may include a display device (1000).

[0034] The imaging device (3000) may be positioned to face the display device (1000) for brightness correction of the display device (1000).

[0035] The imaging device (3000) can capture an image of the display surface of the display device (1000) or an image displayed on the display device (1000), and can provide the captured image, i.e., the image data (PDATA), to the luminance correction device (2000). For example, the imaging device (3000) may include a light-receiving element such as a CCD (Charge-Coupled Device) camera or a CMOS (Complementary MOS) sensor. For example, the imaging device (3000) may provide a single image frame generated by capturing an image of the display surface of the display device (1000) once to the luminance correction device (2000). Also, for example, the imaging device (3000) may not include a light-receiving element itself, but may be connected to an external light-receiving element and configured to receive a single image frame captured once by the external light-receiving element.

[0036] The resolution of the image data (PDATA) may depend on the resolution of the light-receiving element, and the resolution of the image may depend on the resolution of the pixels of the display device (DD). When the resolution of the light-receiving element matches the resolution of the pixels, the coordinates of the image data (PDATA) and the coordinates of the image may correspond to each other in a 1:1 ratio. When the resolution of the light-receiving element is lower than the resolution of the pixels, the coordinates of a single image frame and the coordinates of the image may correspond to each other in a 1:a ratio. Here, a may be a real number greater than 1. When the resolution of the light-receiving element is higher than the resolution of the pixels, the coordinates of the image data (PDATA) and the coordinates of the image may correspond to each other in a b:1 ratio. Here, b may be a real number greater than 1.

[0037] The values ​​of the image data (PDATA) can correspond to the luminance values ​​of the image. For example, if the luminance values ​​of specific coordinates of the image are relatively high, the values ​​of the image data (PDATA) may also be relatively high. In addition, the values ​​of the image data (PDATA) can be determined according to the resolution of the light-receiving element.

[0038] The brightness correction device (2000) can control the operation of the display device (1000) (or display panel). For example, the brightness correction device (2000) can control the display device (1000) so that a specific image is displayed on the display device (1000).

[0039] The brightness correction device (2000) can generate a compensation value (or compensation data) for brightness correction of the display device (1000) based on the image data (PDATA). The brightness correction device (2000) can be implemented as an application processor, etc.

[0040] In the embodiments, the luminance correction device (2000) may acquire blur information based on image data (PDATA) and deblur the image data (PDATA) based on the blur information. Here, the blur information may refer to the degree to which the image is blurred. For example, the blur information may refer to the degree to which a dot pattern in which only one pixel emits light is blurred in the image. The blur information may vary depending on the resolution of the image device (3000), image conditions, etc.

[0041] For example, the luminance correction device (2000) can obtain blur information using first image data for a first image (or dot image) including a dot pattern, and generate deblurred data by deblurring second image data for a second image (or image for measuring luminance deviation) based on the blur information.

[0042] For example, the luminance correction device (2000) can derive a weight matrix for correcting a blurred dot pattern into an ideal dot pattern based on the first imaging data, and apply the weight matrix to the second imaging data to generate deblurring data. Here, the inverse matrix of the weight matrix may correspond to blur information. The deblurring data may have imaging accuracy similar to that of imaging data obtained through a high-resolution imaging device.

[0043] Referring to FIG. 2, the first captured image (IMAGE_P1) may represent an image of a single pixel (PX) (or sub-pixel) within a display device (1000) captured using 3×3 light-receiving elements (e.g., CMOS pixels). That is, in the first captured image (IMAGE_P1), a single pixel (PX) of the display device (1000) corresponds to 3×3 light-receiving elements, and in this case, the magnification ratio (MR) of the capturing device (3000) may be 3 (i.e., MR3). The second captured image (IMAGE_P2) may represent an image of a single pixel (PX) within a display device (1000) captured using 6×6 light-receiving elements. That is, in the second captured image (IMAGE_P2), one pixel (PX) of the display device (1000) corresponds to 6×6 light-receiving elements, and in this case, the magnification ratio of the capturing device (3000) may be 6 (i.e., MR6). The third captured image (IMAGE_P3) may represent an image in which one pixel (PX) within the display device (1000) is captured using 12×12 light-receiving elements. That is, in the third captured image (IMAGE_P3), one pixel (PX) of the display device (1000) corresponds to 12×12 light-receiving elements, and in this case, the magnification ratio of the capturing device (3000) may be 12 (i.e., MR12).

[0044] For example, it can be assumed that a single pixel (PX) includes two light sources (i.e., two light sources that emit light simultaneously in response to the same data signal), and that only one pixel (PX) emits light while pixels adjacent to said pixel (PX) do not emit light. In the third image (IMAGE_P3), each of the two light sources is distinguished, and accordingly, the luminance for a single pixel (PX) (furthermore, the luminance for each of the two light sources) can be accurately measured using the third image (IMAGE_P3). In the second image (IMAGE_P2), the two light sources are not distinguished, but a luminance distribution generally appears within a single pixel area, and the luminance for a single pixel (PX) can be measured relatively accurately using the second image (IMAGE_P2). In the first image (IMAGE_P1), the luminance distribution caused by light emitted from a pixel (PX) extends beyond a single pixel area to adjacent pixel areas (i.e., blurring occurs), and when measuring the luminance of a single pixel (PX) using the first image (IMAGE_P1), the accuracy may be low.

[0045] Referring to FIG. 3, the ratio of interference between the brightness of an emitting pixel and an adjacent pixel is illustrated according to the resolution of the imaging device (3000). The pixel position of the emitting pixel is 0, and the unit of the pixel position may correspond to the number of pixels. For example, a pixel position of 1 may mean the first pixel adjacent to the emitting pixel in a specific direction, and a pixel position of 2 may mean the second pixel adjacent to the emitting pixel in a specific direction.

[0046] When the magnification ratio of the imaging device (3000) is 3 (i.e., MR3), the luminance of the emitting pixel (or image, dot image) may appear to be diffused or blurred by about 57% in adjacent pixels. When the magnification ratio of the imaging device (3000) is 6 (i.e., MR6), the luminance of the emitting pixel may appear to be diffused or blurred by about 17% in adjacent pixels. When the magnification ratio of the imaging device (3000) is 12 (i.e., MR12), the luminance of the emitting pixel may appear to be diffused by about 2% in adjacent pixels.

[0047] That is, the higher the magnification ratio of the imaging device (3000), the more accurately only the luminance of the light-emitting pixel can be measured. However, the higher the magnification ratio of the imaging device (3000), the more expensive the imaging device (3000) may be, or multiple imaging sessions may be required rather than just once to measure the luminance of the entire area of ​​the display device (1000). For example, when the magnification ratio of the imaging device (3000) is 3 (i.e., MR3), the luminance of the entire area of ​​the display device (1000) is measured at once; when the magnification ratio of the imaging device (3000) is 6 (i.e., MR6), the entire area of ​​the display device (1000) is divided into at least 4 areas and the luminance is measured for each; and when the magnification ratio of the imaging device (3000) is 12 (i.e., MR12), the entire area of ​​the display device (1000) is divided into at least 16 areas and the luminance is measured for each. In this case, the imaging time (and the time required for optical compensation therefrom) increases, and the manufacturing cost of the display device (1000) may increase.

[0048] Accordingly, in order to increase the accuracy of the captured image while minimizing the increase in the cost of the optical compensation system (10) (or manufacturing equipment) or the manufacturing cost of the display device (1000), the optical compensation system (10) can perform deblurring on the captured image (or captured data).

[0049] FIG. 4 is a drawing showing an embodiment of a display device included in the optical compensation system of FIG. 1.

[0050] Referring to FIG. 4, the display device (1000) may include a display panel (100), a scanning drive unit (200), a data drive unit (300), a timing control unit (400), and a memory (500).

[0051] The display device (1000) may be implemented as an inorganic light-emitting display device and may include, for example, a flexible display device, a rollable display device, a curved display device, a transparent display device, a mirror display device, etc. As an example, the display device (1000) may be implemented as a display device including a plurality of light-emitting elements having a size of nanoscale to microscale. However, the display device (1000) is not limited thereto, and the display device (1000) may include organic light-emitting elements.

[0052] A display panel (100) includes a plurality of pixels (PX) and can display an image. Specifically, the display panel (100) may include pixels (PX) arranged to be connected to a plurality of scanning lines (SL1 to SLn) and a plurality of data lines (DL1 to DLm). Additionally, the pixels (PX) may be connected to a plurality of sensing control lines (SSL1 to SSLn).

[0053] In one embodiment, each of the pixels (PX) may emit light of one of red, green, and blue. However, as this is exemplary, each of the pixels (PX) may emit light of cyan, magenta, yellow, etc. A first driving voltage (VDD) and a second driving voltage (VSS) may be supplied to the display panel (100) to be applied to a plurality of pixels (PX). According to an embodiment, an initialization voltage (VINT) may be further supplied to the display panel (100) to initialize a plurality of pixels (PX).

[0054] The scanning driver (200) can provide a scanning signal to pixels (PX) of the display panel (100) through scanning lines (SL1 to SLn). The scanning driver (200) can provide a scanning signal to the display panel (100) based on a scanning control signal (SCS) received from the timing control unit (400).

[0055] The data driving unit (300) can provide a data signal in which image data (CDATA) is applied to pixels (PX) of the display panel (100) through data lines (DL1 to DLm). The data driving unit (300) can provide a data signal (or data voltage) to the display panel (100) based on a data driving control signal (DCS) received from the timing control unit (400). In one embodiment, the data driving unit (300) can convert the image data (CDATA) into an analog data signal.

[0056] The timing control unit (400) can receive input image data (IDATA) from an external graphic source (e.g., an application processor) and control the operation of the scanning drive unit (200) and the data drive unit (300) by receiving control signals from the outside. The timing control unit (400) can generate a scanning control signal (SCS) and a data drive control signal (DCS). In one embodiment, the timing control unit (400) can generate image data (CDATA) based on the input image data (IDATA). For example, the timing control unit (400) can convert the input image data (IDATA) into image data (CDATA) to correspond to the arrangement of pixels (PX) within the display panel (100). The image data (CDATA) can be provided to the data drive unit (300).

[0057] In one embodiment, the timing control unit (400) can generate image data (CDATA) by compensating the input image data (IDATA) based on a compensation value (CV) provided from the memory (500). The compensation value (CV) can be generated by a brightness correction device (2000) to compensate for a brightness deviation of the display panel (100) and stored in the memory (500).

[0058] FIG. 5 is a block diagram showing one embodiment of a luminance correction device included in the optical compensation system of FIG. 1. FIG. 6 is a block diagram showing another embodiment of a luminance correction device included in the optical compensation system of FIG. 1.

[0059] First, referring to FIGS. 1 and FIGS. 5, the luminance correction device (2000) may include a luminance measurement block (2100) (or, a resolution conversion block), a deblurring block (2200), and a compensation value calculation block (2300).

[0060] The luminance measurement block (2100) can extract the luminance value of each pixel (PX) based on the image data (PDATA) provided from the image device (3000).

[0061] For example, when the luminance measurement block (2100) receives the first captured image (IMAGE_P1) (or the corresponding captured data) shown in FIG. 2, it can extract the central value among the 3×3 values ​​corresponding to one pixel (PX).

[0062] In other words, the luminance measurement block (2100) can output converted data (WDATA) by converting the resolution of the captured data (PDATA) (or captured image) so that it becomes equal to the resolution of the display device (1000). For example, the luminance measurement block (2100) can convert the resolution of the captured data (PDATA) through image warping.

[0063] Although it has been described that the luminance measurement block (2100) performs image warping on the captured data (PDATA), it is not limited thereto. If the resolution of the captured data (PDATA) is the same as the resolution of the display device (1000), the luminance measurement block (2100) may be omitted.

[0064] The deblurring block (2200) can generate deblurring data (DDATA) by performing deblurring on the converted data (WDATA).

[0065] In the embodiments, the deblurring block (2200) derives blur information or a weight matrix based on the transformed data (WDATA) and can deblur the transformed data (WDATA) based on the blur information or the weight matrix. The deblurring block (2200) will be described later with reference to FIG. 7.

[0066] The compensation value calculation block (2300) can calculate a compensation value or a compensation coefficient for each pixel (PX) based on deblurring data (DDATA). For example, the compensation value may be a grayscale value to be reflected in the input grayscale value for the pixel (PX), and the compensation coefficient may be a coefficient of a compensation function for calculating the compensation grayscale value.

[0067] For example, the compensation value calculation block (2300) can calculate the average luminance value of pixels (PX) based on deblurring data (DDATA) and calculate the compensation value (CV) for each pixel (PX) based on the average luminance and the luminance value of each pixel (PX). For example, the compensation value calculation block (2300) can determine the difference between the average luminance and the luminance value of a pixel (PX) as the compensation value (CV). However, this is exemplary and is not limited thereto, and the compensation value calculation block (2300) can calculate the compensation value (CV) using various known optical compensation techniques or luminance correction (or stain compensation) techniques.

[0068] According to an embodiment, the luminance correction device (2000) further includes a memory device, and the memory device can store information necessary for the operation of the luminance correction device (2000). For example, the memory device can store a compensation value (CV). For example, the memory device can store a weight matrix calculated in the deblurring block (2200) or an intermediate value calculated during the process of deriving the weight matrix.

[0069] Meanwhile, although it has been described that deblurring is performed after image warping, it is not limited thereto. For example, as shown in FIG. 6, a deblurring block (2200_1) deblurs the image data (PDATA) to generate deblurring data (DDATA), a luminance measurement block (2100_1) converts the resolution of the deblurring data (DDATA) to generate converted data (WDATA), and a compensation value calculation block (2300_1) may calculate a compensation value (CV) based on the converted data (WDATA).

[0070] FIG. 7 is a diagram showing an embodiment of a deblurring block included in the luminance correction device of FIG. 5. FIG. 8 is a diagram showing an embodiment of first image data used in the deblurring block of FIG. 7. FIG. 9 and FIG. 10 are diagrams explaining the operation of a weight matrix generator included in the deblurring block of FIG. 7. FIG. 11 is a diagram showing an embodiment of a weight matrix generator included in the deblurring block of FIG. 7. FIG. 12 is a diagram explaining the process of calculating an error in the weight matrix generator of FIG. 11. FIG. 13 is a diagram explaining the operation of an arithmetic unit included in the deblurring block of FIG. 7.

[0071] First, referring to FIGS. 1, 5, and 7, the deblurring block (2200) may include a weight matrix generator (2210) (or a blur information detector) and an operator (2220) (or an integrator). The weight matrix generator (2210) will be described first with reference to FIGS. 7 through 12, and then the operator (2220) will be described with reference to FIGS. 7 and 13.

[0072] The weight matrix generator (2210) can detect blur information or calculate a weight matrix based on the first image data (PDATA1).

[0073] As illustrated in FIG. 8, the first image data (PDATA1) may be an image or data obtained by capturing a display device (1000, see FIG. 1) that displays a dot image (IMAGE_DOT) (or, the first image). The dot image (IMAGE_DOT) may include dot patterns (DOT11~DOT13, DOT21~DOT23) in which only pixels spaced apart from at least two pixels among the pixels (PX) of the display device (1000) emit light.

[0074] For example, a dot image (IMAGE_DOT) includes a plurality of regions divided based on a reference block (BLK), and in each of the regions, dot patterns (DOT11~DOT13, DOT21~DOT23) can be expressed by having only the target pixel among the pixels in the display device emit light and the adjacent pixels among the pixels adjacent to the target pixel not emit light.

[0075] For example, the reference block (BLK) has a size corresponding to 40×40 pixels, and the dot patterns (DOT11~DOT13, DOT21~DOT23) can be spaced apart along the first and second directions (DR1, DR2) with a spacing corresponding to about 40 pixels. For example, the coordinates of the 11th dot pattern (DOT11) may be (1,1), the coordinates of the 12th dot pattern (DOT12) may be (1,41), the coordinates of the 21st dot pattern (DOT21) may be (41,1), and the coordinates of the 22nd dot pattern (DOT12) may be (41,41). The positions of the dot patterns (DOT11~DOT13, DOT21~DOT23) may be pre-set. The spacing between dot patterns (DOT11~DOT13, DOT21~DOT23) and the positions of dot patterns (DOT11~DOT13, DOT21~DOT23) can be varied.

[0076] To express dot patterns (DOT11~DOT13, DOT21~DOT23), the pixel (PX) corresponding to each of the dot patterns (DOT11~DOT13, DOT21~DOT23) may emit light at the maximum grayscale (e.g., 255 grayscale among 0~255 grayscales), but is not limited thereto.

[0077] Meanwhile, if the display device (1000) includes color pixels that emit different colors, the dot image (IMAGE_DOT) is generated by color, and correspondingly, the first image data (PDATA1) can be generated by color. For example, if the display device (1000) includes a red pixel that emits red light, a green pixel that emits green light, and a blue pixel that emits blue light, the dot image (IMAGE_DOT) and the first image data (PDATA1) can each be generated for red, green, and blue, respectively. Accordingly, the operation of deriving the weight matrix and the deblurring operation described later can also be performed by color. However, it is not limited thereto.

[0078] Depending on at least one of the resolution and imaging conditions of the imaging device (3000), the dot patterns (DOT11~DOT13, DOT21~DOT23) are blurred, and the first imaging data (PDATA1) may include blurred dot patterns (DOT_B, see FIG. 9) corresponding to the dot patterns (DOT11~DOT13, DOT21~DOT23). As described with reference to FIG. 5, image warping is applied to the first imaging data (PDATA1), and the resolution of the first imaging data (PDATA1) may be the same as the resolution of the display device (1000).

[0079] Blur information can indicate the degree to which dot patterns (DOT11~DOT13, DOT21~DOT23) (or blood dot pattern (DOT_B)) are blurred.

[0080] In one embodiment, the weight matrix generator (2210) can derive a weight matrix (W) for converting a blood dot pattern (DOT_B) in the first image data (PDATA1) into an ideal dot pattern (DOT_IDEAL).

[0081] Referring to FIG. 9, the blood dot pattern (DOT_B) is included in the first imaging data (PDATA1) and can correspond to each of the dot patterns (DOT11~DOT13, DOT21~DOT23).

[0082] A weight matrix generator (2210) can generate a deblurring dot pattern (DOT_D) (or deblurring dot data (DOT_DDATA)) by convolution operation between a blood dot pattern (DOT_B) (or, first image data (PDATA1) containing the blood dot pattern (DOT_B)) and a weight matrix (W), and can generate an error (ERROR) between the deblurring dot pattern (DOT_D) (or, deblurring dot data (DOT_DDATA)) and an ideal dot pattern (DOT_IDEAL) (or, ideal dot data (DOT_DDATA_IDEAL) containing the ideal dot pattern (DOT_IDEAL). The weight matrix generator (2210) can derive a weight matrix (W) that minimizes the error (ERROR) through machine learning on the blood dot pattern (DOT_B) (or, the first image data (PDATA1)). For example, the weight matrix generator (2210) can derive a weight matrix (W) that minimizes the error (ERROR) by using the Gradient Descent Algorithm (GDA), which is one of the machine learning techniques.

[0083] For example, the weight matrix (W) may have a size of 3×3. However, this is an example, and the size of the weight matrix (W) may vary, for example, the size of the weight matrix (W) may be 5×5, 7×7, etc.

[0084] In the weight matrix (W), the value of the first entry (or first weight) corresponding to the blood dot pattern (DOT_B), i.e., the first entry value, is defined or expressed as P, the value of the second entries (or second weights) located in the first and second directions (DR1, DR2) relative to the first entry, i.e., the second entry value, is defined as Q, and the value of the third entries (or third weights) located in the diagonal direction relative to the first entry, i.e., the third entry value, can be defined as R. When the first, second, and third entries are expressed in relation to pixels, the first entry corresponds to the target pixel representing the blood dot pattern (DOT_B), and the second and third entries correspond to adjacent pixels adjacent to the target pixel.

[0085] The sum of the first, second, and third component values ​​(P, Q, R) can be 1 (i.e., P+4Q+4R=1). When the first component value (P) and the second component value (Q) are determined, the third component value (R) can be calculated automatically (i.e., R=(1-P-4Q) / 4).

[0086] As illustrated in FIG. 10, the weight matrix generator (2210) can repeatedly perform the process of calculating the error (ERROR) and the process of adjusting the first and second component values ​​(P, Q) (and the corresponding third component value (R)). As the number of iterations increases, the error (ERROR) decreases and converges, and the first and second component values ​​(P, Q) (and the third component value (R)) may converge to specific values. That is, a weight matrix (W) in which the error (ERROR) is minimized can be derived.

[0087] In one embodiment, as illustrated in FIG. 11, the weight matrix generator (2210) may include a matrix generator (2211), a convolution operator (2212), and an error calculator (2213).

[0088] The matrix generator (2211) receives a first component value (P) (or, initial first component value) and a second component value (Q) (or, initial second component value), and can generate a weight matrix (W) based on the first and second component values ​​(P, Q).

[0089] The convolution operator (2212) can generate deblurred dot data (DOT_DDATA) by performing a convolution operation between the first image data (PDATA1) and the weight matrix (W). According to an embodiment, the convolution operator (2212) may be omitted, and the deblurred dot data (DOT_DDATA) may be generated in the operator (2220).

[0090] The error calculator (2213) can calculate the error (ERROR) between the deblurred dot data (DOT_DDATA) and the ideal dot data (DOT_DDATA_IDEAL). According to an embodiment, the ideal dot data (DOT_DDATA_IDEAL) may be obtained by binarizing the image data (PDATA1). The error (ERROR) is provided to the matrix generator (2211) or a separate controller and can be used to change or adjust the first and second component values ​​(P, Q).

[0091] In one embodiment, the error calculator (2213) can calculate the mean squared error between deblurring dot data (DOT_DDATA) and ideal dot data (DOT_DDATA_IDEAL).

[0092] Referring to FIG. 12, deblurring dot data (DOT_DDATA) based on a single deblurring dot pattern (DOT_D) can be represented as a block having a size of 7×7. However, the size of the block is not limited to this, and the size of the block can be set to various sizes between the size of the weight matrix (W) (e.g., 3×3) and the size of the reference block (BLK) (e.g., 40×40). Since the weight matrix (W) is intended to define the degree to which the dot pattern is diffused or blurred in the first image data (PDATA1) (i.e., the relative relationship between adjacent values), the error calculator (2213) can normalize the deblurring dot data (DOT_DDATA) in block units. For example, the error calculator (2213) can normalize the deblurring dot data (DOT_DDATA) so that the sum of the values ​​included in a block having a size of 7×7 becomes 1.

[0093] Afterwards, the error calculator (2213) can calculate the mean squared error by performing a difference operation between the deblurring dot data (DOT_DDATA) (or normalized deblurring dot data) and the ideal dot data (DOT_DDATA_IDEAL).

[0094] In one embodiment, the error calculator (2213) can calculate the mean squared error by normalizing the deblurring dot data (DOT_DDATA) for each position of the dot patterns (DOT11~DOT13, DOT21~DOT23, see FIG. 8), and calculate the error (ERROR) by accumulating the mean squared error corresponding to each of the dot patterns (DOT11~DOT13, DOT21~DOT23, see FIG. 8) overall.

[0095] In one embodiment, the error calculator (2213) (or weight matrix generator (2210)) can sequentially adjust the first component value (P) and the second component value (Q).

[0096] For example, the error calculator (2213) can update the first component value (P) using gradient descent. For example, the error calculator (2213) calculates the amount of change in error (ERROR) by changing only the first component value (P) among the first and second component values ​​(P, Q), and when the error (ERROR) decreases, it reduces the learning rate (i.e., the rate of change of error (ERROR)), and can change the first component value (P) based on the learning rate and the amount of change in error (ERROR).

[0097] Subsequently, the error calculator (2213) can update the second component value (Q) using gradient descent. For example, the error calculator (2213) can calculate the amount of change in error (ERROR) by changing only the second component value (Q) among the first and second component values ​​(P, Q), decrease the learning rate when the error (ERROR) decreases, and change the second component value (Q) based on the learning rate and the amount of change in error (ERROR).

[0098] The error calculator (2213) can sequentially repeat the updating of the first component value (P) and the second component value (Q) using gradient descent until the first component value (P) is not changed (or updated). In other words, if the first component value (P) is not changed during the updating process or if the first component value (P) is no longer updated, the error calculator (2213) can stop machine learning and generate a weight matrix (W) (i.e., an optimized weight matrix) based on the final updated first and second component values ​​(P, Q).

[0099] Referring again to FIG. 7, the operator (2220) can generate first deblurred data (DDATA1) by deblurring the second image data (PDATA2) using a weight matrix (W). The second image data (PDATA2) may be image data obtained when all pixels of the display device (1000) (e.g., pixels expressing the same color) emit light to display an image (or a second image). The first deblurred data (DDATA1) may be provided to the compensation value calculation block (2300) as deblurred data (DDATA, see FIG. 5). However, it is not limited thereto.

[0100] For example, the operator (2220) can generate the first deblurring data (DDATA1) by performing a convolution operation between the second image data (DATA2) and the weight matrix (W).

[0101] As illustrated in FIG. 13, the operator (2220) can perform a convolution operation while shifting the weight matrix (W) in pixel units. For example, a value of 131 can be obtained by a convolution operation between the values ​​(98, 101, 103, 105, 120, 109, 103, 118, 115) in the first region (A1) and the weight matrix (W). Thus, the luminance value of each pixel (PX) in the display device (1000) can be obtained more accurately through the first deblurring data (DDATA1).

[0102] As described above, the deblurring block (2200) can derive a weight matrix (W) (or blur information) through machine learning on the first image data (PDATA1) containing dot patterns (or, blood dot patterns), and generate the first deblurring data (DDATA1) (or, deblurring data (DDATA)) by deblurring the second image data (PDATA2) based on the weight matrix (W). Since the first deblurring data (DDATA1) is deblurred on a pixel (PX) basis, it can include more accurate luminance values ​​for each pixel (PX) within the display device (1000). In other words, problems caused by blurring can be improved and image accuracy can be enhanced.

[0103] FIGS. 14 and FIGS. 15 are drawings showing other embodiments of a deblurring block included in the luminance correction device of FIG. 5. FIG. 16 is a drawing explaining the operation of a noise remover included in the deblurring block of FIGS. 14 and FIG. 15. FIG. 16 shows a spatial frequency histogram (or, spatial histogram) for first deblurring data.

[0104] Referring to FIGS. 7, FIGS. 14, and FIGS. 15, the deblurring block (2200) may further include a noise remover (2230) (or partial deblur).

[0105] The noise remover (2230) can generate second deblurring data (DDATA2) by removing noise (artifact) that occurred in the first deblurring data (DDATA1) during the deblurring process. Here, the noise may be an excessively deblurred value. The second deblurring data (DDATA2) can be provided to the compensation value calculation block (2300) as deblurring data (DDATA, see FIG. 5).

[0106] In one embodiment, the noise remover (2230) detects a value (or location thereof) where noise has occurred through spatial frequency analysis of the first deblurred data (DDATA1) and can replace the value with a value in the second image data (PDATA2) (i.e., the original value that has not been deblurred).

[0107] In one embodiment, as shown in FIG. 15, the noise remover (2230) may include a high-frequency analyzer (2231) and a multiplexer (2232).

[0108] The high-frequency analyzer (2231) can perform spatial frequency analysis on the first deblurring data (DDATA1). For example, the high-frequency analyzer (2231) can perform spatial frequency analysis on the first deblurring data (DDATA1) using an average filter.

[0109] Referring to FIG. 16, the spatial frequency histogram for the first deblurred data (DDATA1) can be represented as having a normal distribution (e.g., a Gaussian distribution) based on the mean value (e.g., 0). For example, the high-frequency analyzer (2231) may determine that values ​​included in the first interval (R1) within the standard deviation (e.g., the interval between ±3σ) are normally deblurred, and that values ​​included in the second interval (R2) outside the standard deviation (e.g., the interval exceeding ±3σ) are excessively deblurred. In this case, the high-frequency analyzer (2231) may provide location information for the values ​​(or pixels) included in the second interval (R2) to the multiplexer (2232).

[0110] The multiplexer (2232) receives the first deblurring data (DDATA1) and the second imaging data (PDATA2), and can replace a portion of the first deblurring data (DDATA1) with the value of the second imaging data (PDATA2) based on location information provided by the high-frequency analyzer (2231). For example, with reference to FIG. 13, the first deblurring data (DDATA1) includes a value of 131, and the multiplexer (2232) can replace the value of 131 with the value of 120 (i.e., the original value) included in the second imaging data (PDATA2) based on location information corresponding to the value of 131. For example, the multiplexer (2232) outputs the value of the first deblurring data (DDATA1), but in response to the location information, it can select and output the value of the second imaging data (PDATA2) instead of the value of the first deblurring data (DDATA1).

[0111] Meanwhile, although it has been described that the multiplexer (2232) selects and outputs the value of the second image data (PDATA2) based on location information, it is not limited thereto. For example, the multiplexer (2232) may output a preset value instead of the value of the second image data (PDATA2).

[0112] As described above, the deblurring block (2200) can detect and remove noise through spatial frequency analysis of the first deblurring data (DDATA1). Accordingly, the second deblurring data (DDATA2) may contain more accurate luminance values ​​for each pixel (PX) within the display device (1000). In other words, the imaging accuracy can be further improved.

[0113] FIGS. 17 to 19 are drawings illustrating the effect of deblurring on image data. FIG. 17 shows values ​​(or luminance values) included in a single horizontal line of image data. FIG. 18 shows the Short Range Uniformity (SRU) of a display device (1000) depending on whether luminance correction (or stain compensation) and deblurring are applied. Short Range Uniformity refers to the luminance uniformity between adjacent pixels. FIG. 19 shows images of a single grayscale displayed on the display device (1000).

[0114] Referring to Fig. 17, it can be seen that when deblurring is performed on the image data, adjacent luminance values ​​are clearly distinguished compared to when deblurring is not performed on the image data.

[0115] Referring to FIG. 18, the short-distance uniformity of a display device (1000) displaying a single grayscale image of 16, 47, 63, 95, or 127 is illustrated.

[0116] When luminance correction (or stain compensation) using a compensation value (CV, see FIG. 4 and FIG. 5) is not performed, the short-distance uniformity was found to be within the range of approximately 94.2% to approximately 96.6%. When luminance correction (or stain compensation) is performed using a compensation value (CV) generated based on image data that has not been deblurred, the short-distance uniformity was found to be within the range of approximately 96.7% to approximately 98.4%. Meanwhile, when luminance correction (or stain compensation) is performed using a compensation value (CV) generated based on deblurred image data (i.e., when stain compensation is performed in 1×1 size units), the short-distance uniformity was within the range of approximately 97.4% to approximately 98.8%, which is higher than the short-distance uniformity when deblurring is not performed. In other words, it can be confirmed that short-distance uniformity is improved through deblurring of the image data.

[0117] Referring to FIG. 19, the first display image (IMAGE1), the second display image (IMAGE2), and the third display image (IMAGE3) may each correspond to a specific single grayscale (e.g., a single grayscale of 127). The first display image (IMAGE1) may be an image displayed on a display device (1000) that does not perform a compensation operation using a compensation value (CV). The second display image (IMAGE2) may be an image displayed on a display device (1000) that uses a compensation value (CV) generated based on un-deblurred image data. The third display image (IMAGE3) may be an image displayed on a display device (1000) that uses a compensation value (CV) generated based on deblurred image data. In the case of the first display image (IMAGE1), relatively dark or bright spots may be visible. In the case of the second display image (IMAGE2) and the third display image (IMAGE3), stains are generally not visible, and in particular, it can be confirmed that the third display image (IMAGE3) has a more uniform brightness.

[0118] FIGS. 20a and FIGS. 20b are circuit diagrams illustrating an embodiment of a pixel included in the display device of FIG. 4. For convenience of explanation, FIGS. 20a and FIGS. 20b show a pixel (PX) located in the j-th row (horizontal line) and in the k-th column.

[0119] Referring to FIGS. 20a and 20b, the pixel (PX) may include a light-emitting element (LD), a first transistor (T1, driving transistor), a second transistor (T2), a third transistor (T3), and a storage capacitor (Cst).

[0120] The first electrode (anode or cathode) of the light-emitting element (LD) is connected to the second node (N2), and the second electrode (cathode or anode) is connected to the second driving voltage (VSS) through the second power line (PL2). The light-emitting element (LD) generates light of a predetermined brightness corresponding to the amount of current supplied from the first transistor (T1).

[0121] In one embodiment, a pixel (PX) includes a plurality of light-emitting elements (LDs), and the light-emitting elements (LDs) may be connected in parallel. As illustrated in FIG. 20a, the plurality of light-emitting elements (LDs) may constitute a single light source (EMU) (or light-emitting unit). Some of the plurality of light-emitting elements (LDs) may have mutually different light-emitting characteristics (e.g., current-luminance characteristics), and the number of light-emitting elements (LDs) may vary per pixel (PX). For example, the light-emitting elements (LDs) may be dispersed in a solvent and supplied per pixel (PX) in an inkjet manner, but the light-emitting elements (LDs) may not be evenly dispersed in the solvent. For example, the light-emitting elements (LDs) are connected between a second node (N2) and a second power line (PL2) through a separate alignment process, but the number of light-emitting elements (LDs) that are properly aligned during the alignment process may not be uniform per pixel (PX). Accordingly, a luminance deviation may occur between adjacent pixels (PX). In particular, compared to the case where a pixel (PX) includes only one light-emitting element (LD), the luminance variation may appear relatively large.

[0122] In another embodiment, a plurality of light-emitting elements (LDs) may constitute a plurality of light sources. As shown in FIG. 20b, some of the plurality of light-emitting elements (LDs) are connected in parallel to form a first light source (EMU1), and some of the remainder of the plurality of light-emitting elements (LDs) are connected in parallel to form a second light source (EMU2), and the first light source (EMU1) and the second light source (EMU2) may be connected in series between a second node (N2) and a second power line (PL2). In this case, a third captured image (IMAGE_P3) and / or a second captured image (IMAGE_P2) shown in FIG. 2 may appear.

[0123] The first electrode of the first transistor (T1) is connected to the first driving voltage (VDD) through the first power line (PL1), and the second electrode can be connected to the first electrode of the light-emitting element (LD). The gate electrode of the first transistor (T1) can be connected to the first node (N1). The first transistor (T1) can control the amount of current flowing to the light-emitting element (LD) in response to the voltage of the first node (N1).

[0124] The first electrode of the second transistor (T2) is connected to the data line (DLk), and the second electrode can be connected to the first node (N1). The gate electrode of the second transistor (T2) can be connected to the scan line (SLj). The second transistor (T2) is turned on when a scan signal is supplied to the scan line (SLj) and can transmit a data signal from the data line (DLk) to the first node (N1).

[0125] The third transistor (T3) can be connected between the lead line (RLk) and the second electrode of the first transistor (T1) (i.e., the second node (N2)). That is, the first electrode of the third transistor (T3) can be connected to the lead line (RLk), the second electrode can be connected to the second electrode of the first transistor (T1), and the gate electrode of the third transistor (T3) can be connected to the sensing control line (SSLj). The third transistor (T3) can be turned on when a control signal is supplied to the sensing control line (SSLj) to electrically connect the lead line (RLk) and the second node (N2, i.e., the second electrode of the first transistor (T1)).

[0126] In one embodiment, when the third transistor (T3) is turned on, an initialization voltage (VINT, see FIG. 4) may be supplied to the second node (N2). In another embodiment, when the third transistor (T3) is turned on, the current generated in the first transistor (T1) may be supplied to a sensing unit (not shown).

[0127] A storage capacitor (Cst) can be connected between a first node (N1) and a second node (N2). The storage capacitor (Cst) can store a voltage corresponding to the voltage difference between the first node (N1) and the second node (N2).

[0128] Meanwhile, in an embodiment of the present invention, the circuit structure of the pixel (PX) is not limited by FIG. 20a and FIG. 20b. For example, the light-emitting element (LD) may be located between the first power line (PL1) and the first electrode of the first transistor (T1). Additionally, a parasitic capacitor may be formed between the gate electrode (i.e., the first node (N1)) and the drain electrode of the first transistor (T1).

[0129] Meanwhile, although the transistors (T1, T2, T3) in FIG. 20a and 20b are depicted as NMOS, the present invention is not limited thereto. For example, at least one of the transistors (T1, T2, T3) may be formed as a PMOS. Additionally, the transistors (T1, T2, T3) depicted in FIG. 20a and 20b may be thin-film transistors comprising at least one of an oxide semiconductor, an amorphous silicon semiconductor, and a polycrystalline silicon semiconductor.

[0130] Hereinafter, with reference to FIG. 21, we will examine a light-emitting element according to one embodiment.

[0131] FIG. 21 is a perspective view illustrating a light-emitting element included in the pixel of FIG. 20a.

[0132] Referring to FIG. 21, the light-emitting element (LD) includes a first semiconductor layer (11), a second semiconductor layer (13), and an active layer (12) located between the first semiconductor layer (11) and the second semiconductor layer (13). As an example, the light-emitting element (LD) may be composed of a laminate in which the first semiconductor layer (11), the active layer (12), and the second semiconductor layer (13) are sequentially stacked along the length (L) direction.

[0133] The light-emitting element (LD) may be provided in a rod shape, that is, a cylindrical shape, that extends along one direction. If the extension direction of the light-emitting element (LD) is called the length (L) direction, the light-emitting element (LD) may have one end and the other end along the length (L) direction. Although a columnar light-emitting element (LD) is illustrated in FIG. 21, the type and / or shape of the light-emitting element according to one embodiment is not limited thereto.

[0134] The first semiconductor layer (11) may include at least one n-type semiconductor layer. For example, the first semiconductor layer (11) may include any one of the semiconductor materials InAlGaN, GaN, AlGaN, InGaN, AlN, and InN, and may include an n-type semiconductor layer doped with a first conductive dopant such as Si, Ge, Sn, etc. However, the material constituting the first semiconductor layer (11) is not limited thereto, and the first semiconductor layer (11) may be composed of various other materials.

[0135] The active layer (12) is disposed on the first semiconductor layer (11) and may be formed as a single or multiple quantum well structure. In one embodiment, a clad layer (not shown) doped with a conductive dopant may be formed on the upper and / or lower portions of the active layer (12). For example, the clad layer may be formed as an AlGaN layer or an InAlGaN layer. Depending on the embodiment, materials such as AlGaN and InAlGaN may be used to form the active layer (12), and various other materials may also constitute the active layer (12).

[0136] When a voltage greater than the threshold voltage is applied to both ends of the light-emitting element (LD), electron-hole pairs combine in the active layer (12), causing the light-emitting element (LD) to emit light. By controlling the light emission of the light-emitting element (LD) using this principle, the light-emitting element (LD) can be used as a light source for various light-emitting devices, including pixels of a display device.

[0137] The second semiconductor layer (13) is disposed on the active layer (12) and may include a semiconductor layer of a different type from the first semiconductor layer (11). For example, the second semiconductor layer (13) may include at least one p-type semiconductor layer. For example, the second semiconductor layer (13) may include at least one semiconductor material among InAlGaN, GaN, AlGaN, InGaN, AlN, and InN, and may include a p-type semiconductor layer doped with a second conductive dopant such as Mg, Zn, Ca, Sr, Ba, etc. However, the material constituting the second semiconductor layer (13) is not limited thereto, and various other materials may also constitute the second semiconductor layer (13).

[0138] In the above-described embodiment, the first semiconductor layer (11) and the second semiconductor layer (13) are each described as being composed of a single layer, but the present invention is not limited thereto. In one embodiment, depending on the material of the active layer (12), the first semiconductor layer (11) and the second semiconductor layer (13) may each further include one or more layers, such as a clad layer and / or a TSBR (Tensile Strain Barrier Reducing) layer. The TSBR layer may be a strain relaxation layer that acts as a buffer to reduce the difference in lattice constants by being disposed between semiconductor layers with different lattice structures. The TSBR layer may be composed of a p-type semiconductor layer such as p-GaInP, p-AlInP, p-AlGaInP, etc., but the present invention is not limited thereto.

[0139] Additionally, according to an embodiment, the light-emitting element (LD) may further include an insulating film (14) provided on its surface. The insulating film (14) may be formed on the surface of the light-emitting element (LD) to surround the outer surface of the active layer (12), and may also surround a portion of the first semiconductor layer (11) and the second semiconductor layer (13). However, according to an embodiment, the insulating film (14) may expose both ends of the light-emitting element (LD) having different polarities. For example, the insulating film (14) may expose one end of each of the first semiconductor layer (11) and the second semiconductor layer (13) located at both ends of the light-emitting element (LD) in the length (L) direction, for example, the two bottom surfaces of a cylinder (the upper and lower surfaces of the light-emitting element (LD)) without covering them.

[0140] When an insulating film (14) is provided on the surface of the light-emitting element (LD), particularly on the surface of the active layer (12), it is possible to prevent the active layer (12) from being short-circuited with at least one electrode (for example, at least one contact electrode among the contact electrodes connected to both ends of the light-emitting element (LD)) that is not shown. Accordingly, electrical stability of the light-emitting element (LD) can be secured.

[0141] In addition, by including an insulating film (14) on the surface of the light-emitting element (LD), surface defects of the light-emitting element (LD) can be minimized, thereby improving lifespan and efficiency. Furthermore, if each light-emitting element (LD) includes an insulating film (14), unwanted short circuits between the light-emitting elements (LDs) can be prevented even when multiple light-emitting elements (LDs) are arranged in close proximity to each other.

[0142] In addition, in one embodiment, the light-emitting element (LD) may be manufactured through a surface treatment process. For example, when a plurality of light-emitting elements (LDs) are mixed in a fluid solution (or solvent) and supplied to each light-emitting region (e.g., the light-emitting region of each pixel), each light-emitting element (LD) may be surface-treated so that the light-emitting elements (LDs) are uniformly dispersed within the solution without unevenly aggregating.

[0143] In one embodiment, the light-emitting element (LD) may include additional components in addition to the first semiconductor layer (11), the active layer (12), the second semiconductor layer (13), and the insulating film (14). For example, the light-emitting element (LD) may additionally include one or more phosphor layers, active layers, semiconductor layers, and / or electrodes disposed on one side of the first semiconductor layer (11), the active layer (12), and the second semiconductor layer (13).

[0144] Light-emitting diodes (LDs) can be used in various types of devices requiring a light source, including display devices. For example, at least one light-emitting diode (LD), for instance, a plurality of light-emitting diodes (LDs) each having a size ranging from nanoscale to microscale, can be arranged in each pixel area of ​​a display device, and the light source (or light source unit) of each pixel can be constructed using the light-emitting diodes (LDs). However, the application field of the light-emitting diode (LD) in the present invention is not limited to display devices. For example, the light-emitting diode (LD) can also be used in other types of devices requiring a light source, such as lighting devices.

[0145] FIG. 22 is a flowchart illustrating an optical compensation method according to embodiments of the present invention.

[0146] Referring to FIG. 1, FIG. 7, and FIG. 22, the method of FIG. 22 can be performed in the optical compensation system (10) of FIG. 1.

[0147] The method of FIG. 22 can display a dot image (IMAGE_DOT, see FIG. 8) (or a first image) through a display device (1000) (S100).

[0148] The method of FIG. 22 can generate first image data (PDATA1) by capturing a dot image (IMAGE_DOT) (or a display device (1000) that displays the dot image (IMAGE_DOT)) through an imaging device (3000) (S200).

[0149] As described with reference to FIG. 5, the method of FIG. 22 can convert the resolution of the first image data (PDATA1) to be equal to the resolution of the display device (1000) through image warping.

[0150] The method of FIG. 22 can display a second image through a display device (1000) (S300). As described with reference to FIG. 7, the second image may be an image in which all pixels of the display device (1000) (e.g., pixels expressing the same color) emit light.

[0151] The method of FIG. 22 can generate second image data (PDATA2) by capturing a second image through an imaging device (3000) (S400).

[0152] The method of FIG. 22 can generate deblurred data (DDATA) by deblurring the second image data based on the first image data through a brightness correction device (2000, see FIG. 5) (S500).

[0153] Subsequently, the method of FIG. 22 calculates a compensation value (CV) (or compensation coefficient, compensation data) based on deblurring data (DDATA) through a brightness correction device (2000) (S600), and can store the compensation value (CV) in the memory (500) of the display device (1000) (S700).

[0154] FIG. 23 is a flowchart explaining the process of deblurring the second image data of FIG. 22.

[0155] Referring to FIGS. 7, FIGS. 14, FIGS. 22, and FIGS. 23, the method of FIG. 23 (or image processing method) can derive blur information or a weight matrix (W) based on the first captured data (PDATA1) (S510). Blur information refers to the degree to which the captured image is blurred, and the inverse of the weight matrix can correspond to the blur information. As described with reference to FIG. 7, the method of FIG. 23 can derive the weight matrix (W) through machine learning on the first captured data (PDATA1), for example, through machine learning using gradient descent.

[0156] As described with reference to FIG. 11, the method of FIG. 23 can calculate deblurring dot data (DOT_DDATA) containing a deblurring dot pattern (DOT_D) using a weight matrix (W), calculate an error (ERROR) between the deblurring dot data (DOT_DDATA_IDEAL) containing an ideal dot pattern (DOT_IDEAL) and the deblurring dot data (DOT_DDATA), and adjust the component values ​​(or weights) of the weight matrix (W) based on the error (ERROR). The method of FIG. 23 can repeat the calculation of the error (ERROR) and the adjustment of the component values ​​so that the error (ERROR) is minimized.

[0157] According to the embodiment, the derivation of the weight matrix (W) may be performed after the first image data (PDATA1) is acquired or after the second image data (PDATA2) is acquired.

[0158] Afterwards, the method of FIG. 23 can generate first deblurred data (DDATA1) by deblurring the second image data (PDATA2) based on blur information or a weight matrix (W) (S520).

[0159] Subsequently, the method of FIG. 23 can detect artifacts through spatial frequency analysis of the first deblurring data (DDATA1) and remove the artifacts (S530).

[0160] As explained with reference to FIG. 14, the method of FIG. 23 can replace the noise generated in the first deblurring data (DDATA1) with the value of the second imaging data (PDATA2).

[0161] Figure 24 is a flowchart explaining the process of deriving the weight matrix of Figure 23.

[0162] Referring to FIGS. 11, 23, and 24, the method of FIG. 24 can determine the optimal weight matrix (W) by sequentially and iteratively adjusting the first component value (P) and the second component value (Q) of the weight matrix (W).

[0163] As explained with reference to FIG. 11, the method of FIG. 24 can update the first component value (P) using gradient descent (S511) and update the second component value (Q) using gradient descent (S512).

[0164] The method of FIG. 24 can determine whether the first component value (P) changes (or is updated) (S513). If the first component value (P) changes (e.g., if the first component value (P) is updated to a new value or a value outside the reference range), the method of FIG. 24 can repeat the updating of the first and second component values ​​(P, Q). Conversely, if the first component value (P) does not change, or if the first component value (P) is updated within the reference range (e.g., a range of 1% or less), the method of FIG. 24 can stop the updating (or repetition thereof) of the first and second component values ​​(P, Q). In this case, the method of FIG. 24 can generate a weight matrix (W) using the finally updated first and second component values ​​(P, Q) (S514).

[0165] Although the present invention has been described above with reference to preferred embodiments, those skilled in the art or those with ordinary knowledge in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and technical scope of the invention as described in the claims set forth below.

[0166] Therefore, the technical scope of the present invention should not be limited to the contents described in the detailed description of the specification, but should be determined by the claims. Explanation of the symbols

[0167] 10: Optical compensation system 1000: Display device 2000: Luminance correction device 2100: Luminance measurement block 2200: Deblurring Block 2210: Weight Matrix Generator 2211: Matrix Generator 2212: Convolutional operator 2213: Error Calculator 2220: Arithmetic unit 2230: Noise Canceller 2231: High-frequency analyzer 2232: Multiplexer 2300: Reward value calculation block 3000: Imaging device

Claims

Claim 1 A step of displaying a first image through a display device, wherein the first image includes a dot pattern in which pixels spaced apart with at least two pixels in between emit light; a step of capturing the first image displayed on the display device through an imaging device to generate first image data; a step of displaying a second image through the display device; a step of capturing the second image displayed on the display device through the imaging device to generate second image data; a step of deblurring the second image data based on the first image data to generate deblurred data; and a step of generating compensation data for brightness correction of the display device based on the deblurred data. An optical compensation method comprising the step of storing the compensation data in a memory device, wherein the step of generating the deblurring data includes the step of detecting blur information indicating the degree of blurring of the dot pattern in an image captured by the imaging device based on the first imaging data, and the step of detecting the blur information includes the step of deriving a weight matrix for converting the blurred dot pattern included in the first imaging data into an ideal dot pattern. Claim 2 An optical compensation method according to claim 1, wherein the first image includes a plurality of regions divided based on a reference block, and in each of the regions, a target pixel among the pixels in the display device emits light and adjacent pixels among the pixels adjacent to the target pixel do not emit light, thereby expressing the dot pattern. Claim 3 An optical compensation method according to claim 1, wherein the dot pattern is blurred according to at least one of the imaging conditions and the resolution of the imaging device, and the first imaging data includes a blurred dot pattern corresponding to the dot pattern. Claim 4 An optical compensation method according to claim 1, wherein the step of generating the deblurring data further comprises the step of deblurring the second imaging data based on the blur information. Claim 5 delete Claim 6 An optical compensation method according to claim 1, wherein the step of deriving the weight matrix comprises the step of calculating a first weight for a target pixel corresponding to the dot pattern and a second weight for adjacent pixels adjacent to the target pixel through machine learning on the blurred dot patterns, and wherein the first weight and the second weight are included in the weight matrix. Claim 7 An optical compensation method according to claim 6, wherein a Gradient Descent Algorithm is used in the machine learning. Claim 8 An optical compensation method according to claim 6, wherein the step of deriving the weight matrix comprises: a step of calculating deblurring dot data including a deblurring dot pattern using the weight matrix; a step of calculating an error between ideal dot data including an ideal dot pattern and the deblurring dot data; and a step of adjusting the first weight and the second weight based on the error. Claim 9 An optical compensation method according to claim 8, wherein the step of deriving the weight matrix comprises the step of repeating the step of calculating the error and the step of adjusting the first weight and the second weight so as to minimize the error. Claim 10 In claim 8, the step of calculating the error is an optical compensation method that calculates the error by normalizing the first imaging data. Claim 11 In claim 4, the step of generating the deblurring data further comprises: a step of generating the first deblurring data by deblurring the second imaging data; a step of detecting noise through spatial frequency analysis of the first deblurring data, wherein the noise is a value deblurred outside a reference range during deblurring of the second imaging data; and a step of replacing the noise with a corresponding value of the second imaging data. Claim 12 An optical compensation method according to claim 1, wherein the step of generating the first imaging data includes the step of converting the resolution of an image captured by the imaging device so as to be equal to the resolution of the display device. Claim 13 An optical compensation method according to claim 1, wherein the compensation data stored in the memory device is used to compensate for brightness deviation when driving the display device. Claim 14 An image processing method for preprocessing an image for optical compensation of a display device, comprising: a step of detecting blur information indicating the degree to which a dot pattern is blurred in a first image containing a dot pattern; and a step of deblurring a second image based on the blur information, wherein the step of detecting the blur information includes a step of deriving a weight matrix for converting a blurred dot pattern included in the first image into an ideal dot pattern. Claim 15 delete Claim 16 In claim 14, the step of deriving the weight matrix comprises the step of calculating a first weight for a target pixel corresponding to the dot pattern and a second weight for adjacent pixels adjacent to the target pixel through machine learning on the blurred dot patterns, wherein the first weight and the second weight are included in the weight matrix, an image processing method. Claim 17 An image processing method according to claim 16, wherein gradient descent is used in the machine learning. Claim 18 An image processing method according to claim 16, wherein the step of deriving the weight matrix comprises: generating a deblurring dot image including a deblurring dot pattern using the weight matrix; calculating an error between an ideal dot image including an ideal dot pattern and the deblurring dot image; and adjusting the first weight and the second weight based on the error. Claim 19 In claim 14, the step of deblurring the second captured image further comprises: a step of generating a first deblurred image by deblurring the second captured image; a step of detecting noise through spatial frequency analysis of the first deblurred image, wherein the noise is a value deblurred outside a reference range during deblurring of the second captured image; and a step of replacing the noise with a corresponding value of the second captured image. Claim 20 An optical compensation system comprising: an imaging device that captures an image displayed through a display device and generates imaging data; and a luminance correction device that generates compensation data for luminance correction of the display device based on the imaging data, wherein the luminance correction device detects blur information indicating the degree of blurring of the dot pattern by deriving a weighting matrix for correcting the blurred dot pattern into an ideal dot pattern based on first imaging data including a dot pattern, and deblurs the second imaging data based on the blur information.

Images