Method and apparatus for display correction

Abstract

A method and apparatus for correcting for inter-pixel non-uniformity in a display device is disclosed. The method uses pixel-by-pixel gamma correction to alter the luminance data of the pixels to reduce pixel luminance non-uniformity without sacrificing more display luminance or gray scale dynamic range than is necessary.

Description

Technical Field

[0001] This invention relates to an apparatus and method for display correction. In particular, it is used to correct inter-pixel non-uniformity in display devices comprising multiple LED pixels, such as micro-LED display panels. Background Technology

[0002] III-V semiconductor materials are of particular interest in semiconductor device design, especially III-nitride semiconductor materials.

[0003] "III-V group" semiconductors include binary, ternary, and quaternary alloys of group III elements such as Ga, Al, and In, and group V elements such as N, P, As, and Sb, and are of great interest for several applications, including electronics and optoelectronics.

[0004] Of particular interest are semiconductor materials known as "Group III nitrides," which include gallium nitride (GaN), indium nitride (InN), and aluminum nitride (AlN), as well as their ternary and quaternary alloys. (Al,In)GaN is a term encompassing AlGaN, InGaN, and GaN. Group III nitride materials have not only achieved commercial success in solid-state lighting and power electronics, but also exhibit unique advantages in quantum light sources and light-matter interaction.

[0005] Despite the commercial value of various group III nitride materials, gallium nitride (GaN) is widely regarded as one of the most important new semiconductor materials and is of particular interest for several applications.

[0006] This invention can be implemented using LEDs formed from GaN and InGaN, but is helpfully applicable to LEDs containing alternative group III nitride material combinations. It will be understood that this invention can also be implemented using known LED display technologies.

[0007] Introducing porosity into bulk group III nitrides, such as GaN, is known to profoundly affect their material properties (optical, mechanical, electrical, and thermal). Therefore, the possibility of tuning various material properties of GaN and group III nitride semiconductors by altering their porosity makes porous GaN a promising candidate for optoelectronic applications.

[0008] In a significant improvement over one of the known three-color LED displays, the applicant, Poro Technologies Ltd, has developed variable-wavelength LEDs formed from a group III nitride semiconductor material grown over a porous region of a group III nitride material. Instead of a known LED that emits a single color, these variable-wavelength LEDs can emit a wide range of different emission wavelengths in response to variations in the driving conditions provided by the variable-wavelength LED. These variable-wavelength LEDs and their manufacturing methods are disclosed in International Patent Application No. PCT / GB2022 / 051997, Publication No. WO2023 / 007174.

[0009] Because the peak emission wavelength of such variable-wavelength LEDs can be tuned by changing the driving conditions, a single variable-wavelength LED can replace multiple "monochrome" subpixels. In a particularly preferred embodiment, identical variable-wavelength LEDs can be tuned to emit any of red, green, or blue light by tuning the driving conditions supplied to each other. This means that a color display can be formed from an array of variable-wavelength LEDs, with each variable-wavelength LED acting as a color-variable pixel, or as an alternative to a color-variable subpixel in a subpixelated display.

[0010] Compared to known RGB subpixel displays, the ability of variable wavelength LEDs to emit such a wide range of emission wavelengths makes the display driving and calibration process significantly more complex. Summary of the Invention

[0011] This invention is defined in an independent claim within the scope of the patent application, and reference should now be made to that independent claim. Preferred or helpful features of the invention are defined in the appended claims.

[0012] This invention can be implemented using a display device comprising a plurality of variable wavelength LEDs. In some embodiments, such a display device includes an array of variable wavelength LEDs, and each variable wavelength LED forms a pixel of the display device.

[0013] An example of a variable wavelength LED that can be used in this invention is commonly known as a Dynamic Pixel Tuning (DPT®) variable wavelength LED, which is from Poro Technologies Ltd. and is disclosed in International Patent Application No. PCT / GB2022 / 051997, Publication No. WO2023 / 007174.

[0014] A continuous spectrum of different wavelengths spanning an emission wavelength range can be directly emitted from a variable-wavelength LED by varying the driving conditions supplied to that LED (the amplitude of the driving current and / or driving voltage). In the CIE xy color space, a variable-wavelength LED can emit any color along a continuous wavelength curve. The length and shape of the emitting wavelength lines in the CIE xy color space are determined by the LED composition and structure, as described in WO2023 / 007174.

[0015] One way to control the wavelength of light emitted by a variable-wavelength LED is to provide the LED with a continuously variable drive current. When a display device receives a signal to identify a target color to be displayed by a given variable-wavelength LED, it can then deliver a drive current to that pixel according to the amplitude required for the LED to emit a specific wavelength from its continuous spectrum of emitting wavelengths. However, this approach has disadvantages such as the difficulty in consistently and precisely controlling the variable drive current, and the limitation that variable-wavelength LEDs can only display chromaticity within a continuous range of emission wavelengths that the LED can directly emit.

[0016] One alternative to driving variable-wavelength LEDs is to select a predetermined set of "primary colors" (dominant wavelengths) from the continuous spectrum of emission wavelengths that the variable-wavelength LED can emit. Instead of controlling the emission wavelength to any point across the entire range of emitable colors, a finite set of operating points can be established by selecting multiple N primary wavelengths from the range of emitable colors. Each of the N primary colors is used as a predetermined operating point for the variable-wavelength LED, and the N individual driving conditions required to generate those N primary wavelengths are planned into the display device incorporating the variable-wavelength LED.

[0017] A color palette of N primary colors is formed by selecting N dominant wavelengths from the range of wavelengths that can be emitted by the LED, so that the variable wavelength LED can be controlled at any time to emit light according to one of these N predetermined primary wavelengths. The N primary colors define a color gamut of multiple displayable colors, because any color within this color gamut can be displayed by using a variable wavelength LED by mixing the available primary wavelengths.

[0018] When a display device receives a signal identifying a target color to be displayed by a given variable wavelength LED, the device can select a smaller set of two or more mixable primary colors from a palette of N available primary colors to display the target color. The selection of which primary colors from the N available primary colors depends on the wavelength of the target color to be displayed.

[0019] By driving LEDs to emit only at multiple “primary” wavelengths, rather than at any wavelength from a continuous spectrum of directly emitting wavelengths, display devices can display additional chromaticity that variable-wavelength LEDs cannot directly emit. The primary wavelengths can be mixed using color mixing techniques known to existing displays, so that the overall color observed by a viewer is a spatiotemporal and / or spatial combination of one of the primary colors, rather than a single color “directly” emitted by a single variable-wavelength LED. Driving variable-wavelength LEDs to emit at predetermined “primary” wavelengths also helps simplify driving requirements, as it eliminates the need for a continuously variable drive current; instead, the display device can be calibrated to deliver a more manageable set of N predetermined drive conditions that produce N usable primary colors.

[0020] The number of primary colors "N" available can be selected depending on the desired color gamut size and the desired complexity of the control system. A larger number N results in a larger color gamut that can be displayed, but also requires more driving conditions and more complex control requirements. Since variable wavelength LEDs can emit a continuous range of wavelengths, N can be selected as any number up to infinity (N = infinity is equivalent to continuously driving to any point within the range of emitting wavelengths). N is preferably greater than 3 so that the displayable color gamut is large enough to display a variety of colors. Particularly preferred is N greater than or equal to 4, 5, or 6 to encompass a larger color gamut while keeping control requirements relatively straightforward. Preferably, N can be less than or equal to 8, 10, or 12 to prevent control requirements from becoming overly complex.

[0021] In one embodiment of the display device, each variable wavelength LED is a pixel, and the display device is a field-sequential display. In this embodiment, by driving the LED pixels with discrete driving conditions corresponding to those primary wavelengths, the same variable wavelength LED pixel system is driven to emit multiple discrete primary emission wavelengths in sequential subframes.

[0022] In a field-sequential display, by supplying different driving conditions to LEDs in sequential subframes of a display frame, it is possible to control LEDs of the same variable wavelength to emit multiple discrete primary emission wavelengths one after another. In this way, LEDs of the same variable wavelength can emit one of multiple pixels with a selected "primary emission wavelength" one after another within the duration of a single display frame. During each individual subframe, only one primary emission wavelength is emitted by the pixel. Due to the persistence of vision in the human observer, the primary wavelengths emitted during the sequential subframes are temporally averaged, so that the color observed by a person viewing the display at a normal viewing distance is a mixture of the colors emitted during the display frame.

[0023] In a subpixel display, different primary emission wavelengths can be emitted by individual subpixels. All of these subpixels can be variable-wavelength LEDs, or some subpixels can be non-variable-wavelength LEDs.

[0024] The brightness or grayscale of the emitted light can be altered by shortening or lengthening the duty cycle (pulse width) of the drive current pulses supplied to the LEDs in the display, thus changing the "on time" of the LEDs. In a sequential display, the maximum duty cycle is achieved when drive current is supplied to the variable wavelength LEDs for 100% of the duration of the display frame or subframe assigned to the emitted color. Although controlling the amplitude of the drive current supplied to the variable wavelength LEDs determines the emitted wavelength, the duty cycle of the drive current pulses must still be variable to achieve variable display brightness, adding an additional layer of complexity to device control.

[0025] In Poro Technologies Ltd's Dynamic Pixel Tuning (DPT®) variable wavelength LEDs, the peak emission wavelength of the variable wavelength LED is highly dependent on the amplitude of the electrical drive signal supplied to a given LED. Longer wavelengths of emitted color require lower amplitude drive current / voltage, which inherently results in lower brightness. This poses a challenge for those applications requiring high display brightness, and also presents challenges for color balance when naturally brighter shorter wavelengths must be mixed with naturally darker longer wavelengths. Attached Figure Description

[0026] Embodiments of the present invention will now be described by way of example only, and with reference to the accompanying drawings, wherein:

[0027] Figure 1 This is the CIE xy color space diagram, which shows examples of wavelengths of light that can be emitted by a variable wavelength LED;

[0028] Figure 2 This is a CIE xy color space diagram, which shows an example of superimposing the wavelength of light emitted by a variable wavelength LED with an exemplary point representing the primary color of the emitted wavelength;

[0029] Figure 3 Show Figure 2 The CIE xy color space diagram, where dashed lines show the linear mixing of several pairs of primary colors of emission wavelengths, is used to create a simplified color palette that can display colors;

[0030] Figure 4 This is a diagram illustrating the emission brightness of a pair of uncalibrated emitting pixels according to grayscale levels;

[0031] Figure 5 It is an illustration of three gamma curves showing the relationship between brightness and grayscale levels;

[0032] Figure 6 It is a diagram, its example Figure 4 The emission brightness of the pair of uncalibrated emitting pixels is based on the grayscale level;

[0033] Figure 7 It is a diagram, its example Figure 5 Three gamma curves and two more gamma curves;

[0034] Figure 8 According to a first aspect of the present invention, a method is used to correct inter-pixel non-uniformity in a display device comprising a plurality of LED pixels;

[0035] Figure 9A and Figure 9B This is an example diagram of a dataset of pixel performance used by display devices;

[0036] Figure 10A It is a diagram of normalized brightness across three pixels using grayscale levels;

[0037] Figure 10B yes Figure 10A An illustration of the performance of the three pixels after their brightness levels have been corrected according to the first aspect of the invention;

[0038] Figure 11 This is a diagram illustrating the effect of using black subframes in the correction of pixel brightness data;

[0039] Figure 12A and Figure 12B These are exemplary pixel brightness corrections used in display devices with target gamma values ​​of 1 and 0.5, respectively.

[0040] Figure 13A It is an illustration of the human eye's sensitivity to brightness;

[0041] Figure 13B This is an illustration of a gamma 2.2 curve;

[0042] Figure 14A and Figure 14B These are illustrations of gamma 2.2 and gamma 0.5 curves, respectively.

[0043] Figure 15 According to a second aspect of the invention, there is an illustration of an exemplary hybrid gamma curve;

[0044] Figure 16 According to a second aspect of the invention in the gamma domain, there is an illustration of an exemplary hybrid gamma curve;

[0045] Figure 17 This is a diagram showing the hybrid gamma curve of the second aspect of the invention, which uses the relationship between brightness and grayscale levels as a function and compares it with gamma 2.2 and gamma 1 curves; and

[0046] Figure 18 An example of a method for adjusting the global gamma of a display according to a second aspect of the invention is given. Detailed Implementation

[0047] Figure 1 The CIE xy color space diagram 100 is illustrated, with a continuous black line 101 indicating the wavelength of light that can be emitted by a variable wavelength LED. By changing the driving conditions, such as the magnitude of the driving current or driving voltage supplied to the LED, the wavelength of the light directly emitted by the LED can be tuned to any wavelength along line 101.

[0048] Figure 2 This is the CIE xy color space diagram 200, its example is... Figure 1 The same emission line 101 shown is superimposed with six points representing the "primary colors" P1, P2, P3, P4, P5, and P6. These six primary colors are specific emission wavelengths selected from the continuous emission wavelength line 101. The six individual driving conditions required to supply the variable wavelength LED to generate these six emission wavelengths can be easily found and programmed into the display device. During operation, the six driving conditions can be supplied to the variable wavelength LED, causing the LED to emit light at any of the six primary wavelengths. The space defined by the six primary colors is the color gamut of the variable wavelength LED and the colors that can be displayed at these six primary wavelengths.

[0049] By using field-sequence driving to drive the variable-wavelength LED, different driving conditions can be supplied to the LED during discrete subframe periods. Therefore, during the first subframe period, a first driving condition can be supplied to the LED so that it emits light at a dominant wavelength P1. During a subsequent subframe period, a different driving condition can be supplied to the LED so that it emits light at a different dominant wavelength (one of P2 to P6). In one embodiment, each display frame contains only two subframes, each assigned to emit a different primary color. The overall color observed by a viewer when examining the variable-wavelength LED will be a temporal average of one of the two dominant wavelengths emitted during the display frame period.

[0050] It is possible to divide a display frame into any number of shorter subframes. For example, a display frame could be divided into six subframes so that each of the six primary colors can be emitted one after another within a display frame. However, it is helpful to use only two of the available primary colors during a given display frame, as this simplifies color mixing and reduces power consumption when switching between driving conditions. To reproduce a given target color, two dominant wavelengths from one of the six primary colors P1 to P6 can be selected using a color palette, and then these two selected primary colors are mixed proportionally to give a temporal average of what appears to an observer as the target color.

[0051] like Figure 3As shown, the six available primary colors P1 to P6 can be divided into multiple different pairs. Each pair of primary colors can be linearly mixed with each other to display any color along the straight line 301 between the two primary wavelengths.

[0052] However, from Figure 3 It is also evident that using a finite set of primary colors (e.g., N = 6) means that not all colors in the color gamut lie on the linear blending line 301 between a pair of primary colors. Therefore, because the palette of available primary colors is so small, not all colors in the color gamut can be formed by a single pair of primary colors.

[0053] Display panels formed from variable-wavelength pixels, monochrome (sub)pixels with fixed emission wavelengths, or known pixel technologies will comprise many distinct individual pixels. Micro-LED displays are expected to exhibit inter-emitter variability in intensity, which can result from variations in micro-LED device performance across the wafer, transistor criticality across the display, or other manufacturing tolerances. This variability, occurring inter-emitter (i.e., inter-pixel), causes inter-pixel inhomogeneity across the display; for example, these individual pixels may have different brightness levels when driven by the same drive signal. This variability will result in a seemingly noisy display that cannot accurately reproduce images (whether color or monochrome).

[0054] Therefore, it is desirable to calibrate the display in a way that corrects this inter-pixel variability. Calibrating a micro-LED panel by correcting for micro-LED intensity variations is a process known as "clouding removal." However, this is traditionally corrected for low spatial frequency variations across an LCD panel. Because individual emitters are used instead of a backlight emitter, the required calibration for a micro-LED display is expected to have a very high spatial frequency.

[0055] For a known (i.e., a single-color) microLED, there are two possible methods to compensate for intensity variations: 1) changing the pixel current and 2) changing the pixel duty cycle. Method (1) is difficult because it requires very precise (around microamperes (µA) current control of the microLED on a per-pixel basis. Thus, method (2) is generally preferred because it only requires changing the grayscale value sent to the display.

[0056] For displays composed of monochrome pixels with fixed emission wavelengths, grayscale is typically controlled by pulse-width modulation (PWM). Due to the data loading time associated with process nodes, there is usually a limit of about 8 bits per pixel, meaning that adding dimming capability solely through PWM is not feasible. This means that 8 bits of data typically produce a grayscale range with 256 steps, that is, a relative brightness from 0 (i.e., pixel off) to 255 (i.e., pixel driven at maximum brightness). These steps are called grayscale levels, and the abbreviations GL0 to GL255 will be used herein to refer to these grayscale levels.

[0057] As mentioned above, a pair of pixels in a display can emit different amounts of light when driven by a drive signal corresponding to the GL255. This reduces the accuracy with which the display can reproduce an input image, so the display panel of LED pixels is usually calibrated to overcome this defect. A known method for adjusting the brightness of individual pixels to calibrate the display is to first use a camera to measure the relative intensity non-uniformity of the pixel array in the display. Then a correction item can be evaluated. To linearly compensate for the non-uniformity of the display panel The image to be displayed on the display panel. The following procedure will then be used to process and generate a new image. This new image is sent to the monitor for reproduction:

[0058]

[0059]

[0060] As the panel brightness decreases from its maximum value, the global image brightness is reduced by the least uniform pixels, i.e., according to a factor. To reduce. The factor can be represented by a de-clouding lookup table, which specifies the appropriate grayscale to drive for each individual pixel for any given input grayscale level. The global gamma of the display can then be calibrated by the manufacturer or end user as desired.

[0061] However, intensity compensation via grayscale control has several drawbacks. For example, as discussed, global brightness must be reduced using the least uniform pixels, i.e., according to the factor... This is undesirable and can be particularly problematic for displays intended for use in augmented reality devices / systems, where high brightness levels are crucial for overcoming ambient lighting conditions. Similarly, the grayscale dynamic range is reduced from [0; 255] to [0; The following text will provide a more detailed explanation.

[0062] Figure 4 This is a diagram illustrating the emission brightness of a pair of uncalibrated emitting pixels according to grayscale levels. The diagram has been normalized based on the brightness of pixel 1, thus showing a line 410 for pixel 1 between a brightness of 0 and 1. Pixel 2 is 20% brighter than pixel 1, therefore having a line 420 from a brightness of 0 to 1.2. A dashed line is shown to illustrate the difference in individual brightness that driving pixels 1 and 2 at a grayscale level of GL160 will result in. To make two pixels emit the same brightness, the brighter pixel 2 can operate at a smaller grayscale level compared to pixel 1. The following is driven. However, such grayscale correction effectively overshadows the total brightness output of pixel 2 and reduces the total number of grayscale values ​​between the maximum and minimum (i.e., GL0) brightness of pixel 2, making it more difficult for pixel 2 to provide an accurate representation of an input image pixel.

[0063] Gamma correction

[0064] Gamma describes the relationship between a color value emitted by a display panel and its brightness. For an image in an RGB color space to appear visually correct, the display panel should produce an output brightness that varies linearly with the color values ​​of the input image / pixel. For most displays, gamma correction is required to compensate for the non-linear characteristics of the display panel in order to provide this linearity. Gamma correction first uses a correction function, adapted to the characteristics of the display panel, to adjust the values ​​of the input pixel data before sending it to the display panel itself. This mapping function is typically implemented using a lookup table, and a separate table is usually used for the different color components in an RGB display.

[0065] The three gamma curves are shown in Figure 5 The graph 500 shows the relationship between brightness and grayscale level for gamma curves 510, 520, and 530. Graph 510 is a straight line from 0 to 1, corresponding to a "curve" for gamma 1. Graph 520 is a curve corresponding to gamma values ​​greater than 1, such as gamma 2.2, and graph 530 is a curve corresponding to gamma values ​​less than 1, such as gamma 0.6. As can be seen from graph 500, for the same grayscale level, each of the three different gamma curves will result in a different brightness value. The inventors understand that this characteristic of the gamma curves can be used to calibrate the brightness of one pixel per pixel in a display without using grayscale reduction / proportional adjustments.

[0066] Figure 6 This is diagram 600, its example. Figure 4The emission brightness of the pair of uncalibrated emitting pixels is determined according to the grayscale level. Curves 410 and 420 have been reproduced, and an additional curve 610 has been added to correct one of the exemplary curves 410 using a gamma curve of gamma 0.6. That is, curve 610 shows the brightness of pixel 1 when corrected using the gamma 0.6 curve. As can be seen, the maximum brightness of pixel 1 remains the same, but the brightness initially increases faster with the grayscale level, causing the brightness of pixel 1 at GL160 to match the inherently brighter brightness of pixel 2 at GL160 (and gamma 1). In this way, two pixels can be calibrated to remove non-uniformity at a grayscale level GL160. For a given range of grayscale values, coordinate graphs 420 and 610 are closer than coordinate graphs 420 and 410, and thus... While maintaining a higher average brightness value, using gamma correction results in lower grayscale values ​​than using known grayscale correction across this range. Each pixel also maintains the full range of 256 grayscale levels, thus improving the grayscale reproduction accuracy of pixels compared to known methods.

[0067] Display devices are composed of many different pixels, and thus a variety of different pixel performances / maximum pixel brightness can be expected within the scope of the display panel. In one embodiment, a pixel can be identified as a target pixel, and gamma curve correction can be determined and fine-tuned to maintain the performance of the remaining pixels in the display panel in a manner that better matches the performance of the target pixel. It will be understood that, for a... Pixel storage and processing in a pixel array Individual (and potentially unique) gamma curve corrections can consume significant resources; however, the inventors of this invention understand that the desired gamma curve correction for each pixel can be constructed instead from a reduced set of gamma correction curves, or a combination thereof.

[0068] Figure 7 This is diagram 700, its example. Figure 5 The graph contains three gamma curves: 510, 520, and 530, as well as two more gamma curves: 710 and 720. Coordinate graph 710 has been defined as... The combination of these, and coordinate graph 720 has been defined as The combination of these curves. In this way, three curves, GammaA, GammaB, and GammaC, have been used to generate a set of five different gamma curves that can be used for gamma calibration. It will be understood that, depending on available resources and specific implementation aspects, a desired number of individual gamma curves can be selected, and the resulting gamma combinations can be identified. Generally, the gamma curve used for the first pixel in a display array can be defined as:

[0069]

[0070] in Although the above examples illustrate combinations of three gamma curves, it will be understood that different numbers of gamma curves can be selected for each combination, and in practice, individual gamma curves can be weighted in unequal or equal ways.

[0071] The inventors of this invention understand that an effective component for combining multiple different individual gamma curves for a pixel in a display is to divide the pixel output frame into multiple subframes, and each individual gamma curve is used to correct the input grayscale level of the pixel for that subframe during that subframe. In this way, gamma correction according to coordinate graph 720 can be achieved by dividing the frame into three subframes, correcting the pixel's input grayscale level using a GammaB curve during the first subframe, and then correcting the pixel's input grayscale level using a GammaC curve during the second and third subframes. When the pixel frame is observed by the human eye, the three subframes will be averaged in the time domain to form a single frame with combined gamma correction, which can be called a multi-gamma luminance integration. It will be understood that, for some pixels, if appropriate, the gamma correction curve can be represented by the same individual / component gamma curve in each of the multiple subframes.

[0072] To identify the appropriate gamma curve combination to correct individual pixels, an algorithm can be executed to potentially solve for the gamma combination that provides the best fit for that pixel across the grayscale range, based on one or more boundary conditions. For example, if certain ranges of the grayscale spectrum are expected to be most frequently used in a particular application, the fitting algorithm can be tuned to prioritize these ranges rather than attempting to fit across the entire grayscale spectrum / range. Similarly, the allowable percentage of brightness deviation for each grayscale level can be defined as one of the boundary conditions used by the fitting algorithm.

[0073] Figure 8 Method 800 is illustrated for correcting inter-pixel non-uniformity in a display device comprising a plurality of LED pixels. In a first step 810, luminance data of each pixel of the plurality of LED pixels in the display device when driven by one or more common drive signals is received. This luminance data for each pixel can be acquired using a high-resolution camera by imaging / capturing an image of the plurality of LED pixels in the display device. When the pixels are driven at a single common gray level, for example, the maximum luminance of each pixel is GL255, an image can be captured, or multiple images can be captured in separate steps along the gray range. In one example, images of the pixels are captured at each gray level, such that the luminance data includes the luminance of each pixel at each gray level.

[0074] In step 820, a target brightness characteristic is identified based on the brightness data, wherein the brightness characteristic defines the brightness of one of the target pixels among the plurality of LED pixels when driven by the one or more common drive signals. In known methods, this target pixel would already be the pixel with the lowest brightness output / performance; however, in the first aspect of the invention, the target pixel is preferably a pixel with better performance than the pixel with the lowest brightness output. In one example, the brightness characteristic may correspond to the brightness of the group of pixels with the most common brightness data, i.e., the largest group of pixels with similar brightness data / brightness performance across the grayscale range. This is helpful because it minimizes the number of pixels that need correction, while also avoiding known techniques that scale down to the lowest performance common denominator.

[0075] An example illustration of the pixel performance dataset is shown in Figure 9A In the middle, and the corresponding normalized distribution of pixel performance is based on Figure 9B Example in. Figure 9B The dashed lines in the middle indicate the largest group of pixels with similar brightness performance, which are selected as the "target pixels". The remaining pixels are adjusted to converge as closely as possible to the performance of these target pixels.

[0076] In step 830, gamma correction is determined for each pixel whose luminance data does not match the target luminance characteristics. The gamma correction is configured to change the luminance data for individual pixels (which are not target pixels) to reduce the inhomogeneity between their luminance data and the target luminance characteristics.

[0077] In one example, the brightness data of a pixel is normalized to the brightness data of the target pixel, or a group of target pixels, and can be configured to iteratively refine the combination of gamma curves for each pixel until a desired coordinate map of pixel brightness versus grayscale level is achieved. The critical threshold for iterative refinement can be controlled by the compensation criteria / boundary conditions of the refinement algorithm.

[0078] In one example, a criterion could define a grayscale range within which the fitting algorithm is configured to operate as a boundary condition; that is, the fitting criticality applies only to brightness values ​​within that grayscale range. This can be useful, for instance, if in a particular use case the lower brightness portions of the grayscale range are of limited importance due to higher levels of ambient light, making the range difficult to discern. Alternatively or additionally, a criterion could define the maximum acceptable percentage difference between the brightness of a target pixel at a given grayscale and the brightness of the pixel (for which gamma correction is being calculated) at that given grayscale level.

[0079] In this manner, the fitting algorithm divides the data frames of pixels whose brightness data does not match the target brightness characteristics into multiple subframes, and repeatedly refines the gamma curve for each subframe until one or more criteria are satisfied by the generated gamma correction. Then, in step 840, individual gamma corrections are stored for each pixel so that they can be later used on the input image pixel data to be displayed by the display device. This can be stored in a lookup table for the display device, which may be called a cloud removal factor table.

[0080] The above method defines a standard luminance performance level, which is higher than the performance of pixels with the minimum luminance level. Pixels with performance levels lower than the standard luminance performance level undergo gamma correction to improve their luminance in the mid-to-high grayscale range, thereby converging their performance in this range with the standard performance level. In this way, compared to known cloud removal techniques, the overall brightness of the display is maintained, and a large number of pixels in the display device avoid known reductions in the grayscale range. However, for pixels with a performance level higher than (which can be a certain percentage higher, for example, 10% in one example) the standard luminance performance level, some degree of grayscale reduction may still be required.

[0081] Therefore, for a pixel with higher performance within this range, a grayscale compensation value can be defined and stored in a lookup table for removing cloudiness in the display device. This grayscale compensation value can be an absolute reduction in grayscale or a percentage reduction used to proportionally adjust the grayscale level. Thus, for pixels within this performance range, the grayscale level of the pixel output can be adjusted by gamma correction (e.g., across multiple subframes), and the grayscale level of each frame / subframe is reduced according to the grayscale compensation value.

[0082] Figure 10A This is a graph illustrating the normalized brightness of three pixels across the grayscale level. The pixel represented by curve 1010 is the target pixel with a maximum brightness L; the pixel represented by curve 1020 has higher efficiency and a maximum brightness of 1.2L; and the pixel represented by curve 1030 has lower efficiency and a maximum brightness of 0.9L. In this example, the display is globally calibrated to gamma 2.2, as can be seen from the individual curves. Figure 10B yes Figure 10AAn illustration shows the performance of three pixels after their brightness levels have been corrected according to the first aspect of the invention. Specifically, a pixel with a maximum brightness of 1.2L is represented by curve 1025, and a pixel with a maximum brightness of 0.9L is represented by curve 1035. As can be seen from the results, the performance of the 0.9L and 1.2L pixels converges with the performance of the target pixel across most of the grayscale range. In the highest grayscale range, the performance continues to diverge to a certain extent because the maximum brightness of the 0.9L pixel cannot be increased. However, it can be seen that the maximum brightness of the 1.2L pixel has decreased to approximately 1.1L, for example, by grayscale correction using grayscale compensation values ​​as described above.

[0083] In the example of implementing gamma correction using three subframes, this grayscale compensation value can be applied to pixels with a performance level approximately 10% to 30% higher than a defined standard luminance performance level. In this example, for any pixel with a performance level more than 30% higher than a defined standard luminance performance level, it is expected that the pixel will be turned off in one of the subframes. It will be understood that this will result in a 33% reduction in the pixel's luminance output. For example, a pixel with an inherent maximum luminance of 1.35L compared to the target pixel will then have a maximum luminance output of 0.9L. This is illustrated in... Figure 11 In the diagram, a target pixel is represented by curve 1110 and a high-performance pixel is represented by curve 1120, showing a normalized brightness across the gray level. Curve 1125 represents a high-performance pixel implemented by three subframes, one of which is a black / blank subframe.

[0084] It will be understood that, for any given implementation of the invention described above, the use of a blank / black subframe (e.g., during which the pixels in question are driven at a grayscale level GL0 and thus turned off) can become desirable, depending on the criticality at which the frame is divided into subframes. In one example, pixels can be turned off in this way by selecting a component gamma curve with gamma 0 for that subframe.

[0085] In the above explanation, gamma correction has been discussed for individual pixels. It will be understood that in some practical applications, it may be desirable to apply the resulting gamma correction to a group of neighboring pixels in a display device, such as a pixel block, based on its averaged brightness data. While this may reduce the accuracy of the correction / calibration, it will, for example, reduce the amount of processing required and is applicable to practices with very high pixel densities.

[0086] While the above explanation illustrates how gamma curves are used at the pixel level to correct inter-pixel inhomogeneities, it will be understood that display devices are typically calibrated at a global level using a global target gamma curve (which may differ for each color channel in a color display). In this use case, the target pixels will be calibrated using the target gamma curve, and the processing of other pixels in the display device will converge to the performance of the target pixels processed using the target gamma curve. Thus, the gamma correction determined for pixels with luminance data that does not match the target luminance characteristics will be based separately on the target gamma curve used by the display device.

[0087] In some of the examples above, one of the gamma 2.2 gammas has been exemplified. Figure 12A A diagram illustrating normalized brightness across gray levels for a group of pixels that have been globally calibrated to a target of Gamma 1. Line 1210 represents the target pixel and has a maximum brightness L. Line 1220 represents a pixel with lower efficiency, having a maximum brightness of 0.9L after processing using the aforementioned subframe gamma correction method. Line 1230 represents a pixel with higher efficiency, having a maximum brightness of 1.2L after gray level reduction using gray level compensation values ​​and processing using the aforementioned subframe gamma correction method. As can be seen, the first aspect of the invention also helps to reduce inter-pixel non-uniformity across multiple gray levels for Gamma 1 output.

[0088] Figure 12B A diagram illustrating normalized brightness across gray levels is provided for a group of pixels that have been globally calibrated to a target of gamma 0.5. Line 1250 represents the target pixel and has a maximum brightness L. Line 1260 represents a pixel with lower performance, having a maximum brightness of 0.9L after processing using the aforementioned subframe gamma correction method. Line 1270 represents a pixel with higher performance, having a maximum brightness of 1.1L after gray level reduction using gray level compensation values ​​and processing using the aforementioned subframe gamma correction method. As can be seen, the first aspect of the invention also helps to reduce inter-pixel non-uniformity across multiple gray levels for a gamma 0.5 output.

[0089] Global Gamma Configuration

[0090] The following discussion focuses on the overall control target gamma for the display, not pixel non-uniformity.

[0091] Display devices are typically calibrated to gamma 2.2, which is due to the relationship between the actual brightness and the brightness perceived / sensed by a person's eye. Figure 13AA coordinate graph 1300 illustrating the variation in brightness perceived by a human eye, using the true brightness of a light source as a function, is shown. Region 1310 of the coordinate graph 1300 is highlighted, where the curve has a steep gradient. This region 1310 represents the high sensitivity of the human eye to relative brightness changes at low brightness levels. This means that very small changes in the brightness of an object can be perceived by an observer within this region 1310. In contrast, at higher brightness levels, the slope decreases, and the increase in brightness change perceived by the observer relative to the true brightness will be much greater.

[0092] As mentioned above, grayscale levels for a display are transmitted using a limited amount of data for each pixel, typically 8 bits, to provide 256 levels of dimming, from 0 to 255. To use this data most effectively, the actual brightness increment between grayscale levels is configured to be smaller at lower brightness levels and larger at higher brightness levels. This is achieved by using a gamma curve, such as gamma 2.2. Figure 13B The coordinates are shown in Figure 1350.

[0093] The use of gamma 2.2 in display devices presupposes that the display device is the primary source of brightness observed by the user, and this is validly assumed to be done in a relatively dim environment. However, in some use cases of display technology, such as augmented reality devices, for a significant duration of use, the user may be in a bright environment, such as outdoors in sunlight. In such bright environments, the user's vision will adjust based on ambient lighting, and the user will find it difficult to perceive subtle steps in brightness at the lower levels typically targeted by gamma 2.2 correction.

[0094] In such use cases, the display's brightness will need to exceed a minimum critical brightness level to match the ambient light level, and image details in areas where the display's brightness is below this critical level will be lost or difficult for the user to perceive. For outdoor applications, this minimum brightness critical level can be around 1 kilonits; however, it will be understood that this will depend on the weather conditions.

[0095] In augmented reality applications, the light emitted by the display's light source is first guided through a waveguide and then reaches the surface on which the user views the content. Assuming that the efficiency of this optical system is about 1%, then for a display with a light source of about 800,000 to 3 million nits, the brightness available for viewing by the user will be about 8k to 30k nits (depending on the original brightness of the light source).

[0096] Taking an 8k nit maximum brightness, which can be inspected by a user, as an example, the lowest 12.5% ​​of the brightness produced by the display will be below the 1k nit threshold and therefore difficult for the user to perceive. If the display is calibrated using gamma 2.2 and an 8-bit color depth, this would correspond to 100 grayscale levels that are difficult to perceive, meaning that only about 61% of the grayscale range is usable. Although this can be mitigated by using a brighter display, this example of a 30k nit maximum brightness display would still actually result in only about 78.5% of the grayscale range being usable (the lowest 3.3% of the brightness produced by the display would be below the 1k nit threshold, and this would correspond to 55 grayscale levels).

[0097] As can be seen from the above, although the traditional gamma 2.2 calibration of the display matches human vision well under low brightness conditions, it still suffers severe grayscale loss under high ambient brightness conditions, such as when using augmented reality devices outdoors. Severe grayscale loss in this context means that a large number of grayscale levels at the lower levels are completely imperceptible to the user due to the significant amount of ambient light.

[0098] Using a parallel approach, Figure 14A An example of a gamma 2.2 curve is shown, and Figure 14B An example of a gamma 0.5 curve is shown. Figure 14A The square 1410 in the middle highlights the gamma 2.2 in the low slope section of the low gray level, as mentioned above. This causes a large number of gray levels in this range to drop below the minimum brightness threshold and become difficult for a user to perceive.

[0099] The gamma 0.5 curve has the advantage of rapidly increasing brightness levels at the lower end of the grayscale range, meaning fewer grayscale levels will be used before exceeding the minimum brightness threshold. However, as shown in block 1420, the gamma 0.5 curve increases brightness very slowly at the higher end of the grayscale range. Since the human eye perceives very small changes in brightness across this grayscale range, it is unlikely that a user would be able to distinguish between several adjacent grayscale levels within this high brightness range, thus making this an inefficient use of available grayscale.

[0100] As is understood by the inventors of this invention, more efficient use of one of the 8-bit depths can be achieved by using a hybrid gamma curve. In one embodiment, the hybrid gamma curve may be designed to have a greater brightness / brightness increase at the lower end of the grayscale range per grayscale level (or in other words, a faster increase in a brightness level), while the gamma curve is designed to be higher than a given minimum brightness (i.e., the range of brightness that a user is expected to perceive in an outdoor environment, which may be referred to as the effective brightness range) in a manner similar to a conventional gamma curve (such as gamma 2.2).

[0101] exist Figure 15As illustrated in the diagram, in the exemplary curve 1510, a display system with a maximum user-viewable brightness of 8k nits will exceed the 1k nit threshold at GL36, meaning approximately 86% of the grayscale range will be available, a significant increase compared to the 61% corresponding to gamma 2.2. This results in the effective brightness range being represented by the brightness range shown in the dashed box 1520. In the example of a display system with a maximum user-viewable brightness of 30k nits, this will increase to 98.8% of the usable grayscale range, as it will exceed the 1k nit threshold precisely at GL3. This significantly increases the granularity of grayscale control within the effective brightness range. Furthermore, this means that compared to using a known constant gamma curve, such as gamma 2.2, an improved level of detail / quality can be provided in both the low and high grayscale ranges.

[0102] although Figure 15 The curve 1510 is referred to here as a hybrid gamma curve. It will be understood that it is not a traditional gamma curve because the gamma value associated with each gray level changes as the gray level progresses. Figure 16 A diagram is shown, which is illustrated in the gamma domain. Figure 15 Curve 1510, that is, curve 1510 is expressed based on the equivalent gamma value of the points on the curve as a function of gray level. As can be seen, this exemplary mixed gamma curve 1510 changes from a gamma 0.7 curve at low gray levels to a gamma 2.5 curve as the gray level gradually increases to high gray levels.

[0103] Figure 16 The diagram is divided into three segments or regions: a low grayscale region (e.g., grayscale levels GL1 to GL32), a medium grayscale region (e.g., grayscale levels GL32 to GL192), and a high grayscale region (e.g., grayscale levels GL192 to GL255). As illustrated in the figure by illustrative curve 1510, the rate of change of the gamma value stabilizes at approximately GL5 in the low grayscale region and increases with a slope of approximately 0.0097. In the medium grayscale region, this rate of increase slows to a slope of approximately 0.0073, and then in the high grayscale region, this rate of increase again slows to a slope of approximately 0.0051. The inventors have found that this type of hybrid gamma curve is particularly effective as a global gamma correction for display devices intended for use in bright environments, such as augmented reality display systems.

[0104] In display devices, this hybrid gamma curve can be achieved by splitting one of the sampling frames of luminance data into subframes and applying different gamma curves to each subframe, as discussed above regarding gamma correction methods. In the example where this method splits a frame into three subframes, the hybrid gamma curve can be expressed as:

[0105]

[0106] Alternatively, a hybrid gamma curve can be achieved using high-resolution (e.g., more than 16 bits) PWM output control. In both cases, the result is an effective gamma change in the hybrid curve as the grayscale level changes from GL0 to GL255 (rather than a known gamma curve remaining at a constant gamma value). Figure 17 This is a diagram showing a hybrid gamma curve 1510, which uses the relationship between brightness and grayscale level as a function and compares it to a gamma 2.2 curve 1720 and a gamma 1 curve 1730. As can be seen, the behavior of the hybrid gamma curve 1510 is very close to that of the gamma 2.2 curve 1720 at higher grayscale levels, but provides enhanced brightness for detail in lower and medium grayscale ranges. In this way, the hybrid gamma curve of the present invention provides more detail in the low grayscale range than a known gamma 2.2 curve, and also provides more detail in the high grayscale range than a known gamma 0.5 curve or a known gamma 1 curve.

[0107] Figure 18 According to a second aspect of the present invention, a method 1800 for controlling the brightness of an image signal is illustrated. In a first step 1810, a brightness control member receives image signal brightness data. Next, in step 1820, the brightness control member processes the image signal brightness data using a curve for gamma correction, the curve having a gamma value that varies according to grayscale, and the gamma value increasing as grayscale increases. Next, in step 1830, the brightness control member outputs the image brightness data to a display device.

Claims

1. A method for correcting inter-pixel non-uniformity in a display device comprising a plurality of LED pixels, the method comprising: Receive brightness data of each of the plurality of LED pixels in the display device when it is driven by one or more common driving signals; Based on the brightness data, the target brightness characteristics are identified, which define the brightness of the target pixel among the plurality of LED pixels when driven by the one or more common driving signals. For each pixel whose luminance data does not match the target luminance characteristics, gamma correction is determined. Each gamma correction is configured to change the luminance data of the individual pixel to reduce the inhomogeneity between the luminance data of the individual pixel and the target luminance characteristics. as well as Individual gamma corrections are stored for each pixel to be used in operations on the input image pixel data to be displayed by the display device.

2. The method according to claim 1, wherein, Each established gamma correction is formed by a combination of multiple component gamma curves.

3. The method according to claim 2, wherein, Each of the plurality of component gamma curves is used to operate on the input image pixel data during individual subframes in a plurality of subframes to provide the gamma correction, the plurality of subframes representing a frame in the period during which the input image pixel data is to be displayed by the pixel of the display device.

4. The method according to claim 3, wherein, Each subframe lasts for an equal period of time.

5. The method according to claim 3 or 4, wherein, Each frame contains three subframes.

6. The method according to any one of claims 2 to 5, wherein, The multiple component gamma curves are selected from a set of n gamma curves, where n is less than the number of pixels.

7. The method according to any one of claims 1 to 6, wherein, The target pixel was selected based on the target pixel with the most common brightness data.

8. The method according to any one of claims 1 to 7, wherein, For a pixel having luminance data that is a first percentage range higher than the target luminance characteristic, the method further includes determining and storing a grayscale compensation value for use in input image pixel data to reduce the grayscale level of the pixel according to the grayscale compensation value.

9. The method according to any one of claims 3 to 8, wherein, For a pixel whose luminance data is a second percentage higher than the target luminance characteristic, the method causes that individual pixel to be turned off for one or more of the multiple subframes.

10. The method according to claim 9, wherein, This second percentage is based on the number of subframes that represent one frame.

11. The method according to claim 9 or 10, wherein, This individual pixel is turned off by selecting a component gamma curve of gamma 0 for the subframe.

12. The method according to any one of claims 1 to 11, wherein, The gamma correction used for each pixel whose brightness data does not match the target brightness characteristics is determined by a fitting algorithm based on one or more criteria.

13. The method according to claim 12, wherein, The one or more criteria define the boundary condition gray range in which the fitting algorithm is configured to operate.

14. The method according to claim 12 or 13, wherein, The one or more criteria define the maximum acceptable percentage difference between the brightness of the target pixel at a given gray level and the brightness of the pixel under consideration at that given gray level.

15. The method according to any one of claims 12 to 14, wherein, The fitting algorithm repeatedly refines the gamma correction for each pixel whose brightness data does not match the target brightness characteristics until each of the one or more criteria is met.

16. The method according to any one of claims 1 to 15, wherein, The gamma correction determined for each pixel whose luminance data does not match the target luminance characteristics is based on the target gamma curve of that pixel in the display device.

17. The method according to any one of claims 1 to 16, wherein, Storing the individual gamma correction for each pixel includes storing a lookup table for each pixel.

18. The method according to any one of claims 1 to 17, wherein, A defined gamma correction can be applied to a set of adjacent pixels based on the average brightness data of that set of adjacent pixels on the display device.

19. The method according to any one of claims 1 to 18, wherein, The brightness data used by each of the plurality of LED pixels in the display device is received from a high-resolution camera configured to image the plurality of LED pixels in the display device.

20. The method according to any one of claims 1 to 19, wherein, These multiple LED pixels are multiple variable wavelength LED pixels.

21. A display device comprising a plurality of LED pixels and a processing unit configured to correct inter-pixel non-uniformity by: Retrieve the individually stored gamma correction for each of the plurality of LED pixels in the display device; Receive input image data including input image pixel data for each pixel; as well as The input image pixel data is manipulated using the gamma correction retrieved for individual pixels; The individual gamma correction for each pixel is determined for pixels whose luminance data does not match the target luminance characteristics. Each gamma correction is configured to change the luminance data of the individual pixel to reduce the inhomogeneity between the luminance data of the individual pixel and the target luminance characteristics when driven by one or more common drive signals. as well as The brightness characteristic defines the brightness of a target pixel among the plurality of LED pixels when it is driven by one or more common driving signals.

22. The display device according to claim 21, wherein, Each established gamma correction is formed by a combination of multiple component gamma curves.

23. The display device according to claim 22, wherein, Each of the plurality of component gamma curves is configured to operate on the input image pixel data during individual subframes of a plurality of subframes to provide the gamma correction, the plurality of subframes representing a frame within the period during which the input image pixel data is to be displayed by the pixel of the display device.

24. The display device according to claim 23, wherein, Each subframe lasts for an equal period of time.

25. The display device according to claim 23 or 24, wherein, Each frame contains three subframes.

26. The display device according to any one of claims 22 to 25, wherein, The multiple component gamma curves are selected from a set of n gamma curves, where n is less than the number of pixels.

27. The display device according to any one of claims 21 to 26, wherein, The target pixel is configured to be selected based on the target pixel with the most common brightness data.

28. The display device according to any one of claims 21 to 27, wherein, For a pixel whose brightness data is a first percentage range higher than the target brightness characteristic, the gray level of the input image pixel data for that pixel is configured to be reduced according to the gray level compensation value.

29. The display device according to any one of claims 23 to 28, wherein, For a pixel whose luminance data is a second percentage higher than the target luminance characteristic, the display device is configured to cause that individual pixel to be turned off for one or more of the plurality of subframes.

30. The display device according to claim 29, wherein, This second percentage is based on the number of subframes representing a frame.

31. The display device according to claim 29 or 30, wherein, The individual pixel is configured to be turned off by selecting a component gamma curve of gamma 0 for the subframe.

32. The display device according to any one of claims 21 to 31, wherein the gamma correction for each pixel having luminance data that does not conform to the target luminance characteristics is configured to be determined by a fitting algorithm based on one or more criteria.

33. The display device according to claim 32, wherein, The one or more criteria define the boundary condition gray range in which the fitting algorithm is configured to operate.

34. The display device according to claim 32 or 33, wherein, The one or more criteria define the maximum acceptable percentage difference between the brightness of the target pixel at a given gray level and the brightness of the pixel under consideration at that given gray level.

35. The display device according to any one of claims 32 to 34, wherein, The fitting algorithm repeatedly refines the gamma correction for each pixel whose brightness data does not match the target brightness characteristics until each of the one or more criteria is met.

36. The display device according to any one of claims 21 to 35, wherein, The gamma correction determined for each pixel whose luminance data does not match the target luminance characteristics is based on the target gamma curve of that pixel in the display device.

37. The display device according to any one of claims 21 to 36, wherein, Storing the individual gamma correction for each pixel includes storing a lookup table for each pixel.

38. The display device according to any one of claims 21 to 37, wherein, A defined gamma correction can be applied to a set of adjacent pixels based on the average brightness data of that set of adjacent pixels on the display device.

39. The display device according to any one of claims 21 to 38, wherein, The brightness data used by each of the plurality of LED pixels in the display device is received from a high-resolution camera configured to image the plurality of LED pixels in the display device.

40. The display device according to any one of claims 21 to 39, wherein, These multiple LED pixels are multiple variable wavelength LED pixels.

Images