Method for determining the onset of relaxation after an image burn-in process on pixel-by-pixel optical display devices

Description

[0001] The invention relates to a method for determining the onset of relaxation on a pixel-by-pixel controllable optical display device that has been subjected to a burn-in process. The invention further relates to a method and a device for determining the burn-in behavior on such a display device, as well as the application of such a method for determining the burn-in behavior of displays used in vehicles.

[0002] Electronically controlled pixel-by-pixel optical display devices, especially displays, can exhibit a change in the displayed image known as burn-in. Burn-in can be caused by the prolonged, unchanged display of an image on the display device. If the display device is controlled with a modified display image immediately after displaying such a burned-in image, the displayed image can exhibit features and structures of the previously displayed burn-in image. Such burn-in effects are also referred to as ghosting, image sticking, image retention, or burn-in.

[0003] After burn-in, the display image appears different from the same image on a display device that was not subjected to burn-in, whereby the difference may decrease with the duration of the display image. In many cases, the time course of the disturbance S(t) caused by burn-in can be described by a decaying exponential function. S t = S 0 ⋅ e − t τ The disturbance S 0 , which is observed immediately after the display device switches from the burn-in image to the display image, can be determined, for example, by a scalar factor α to a reference image displayed on the display device after burn-in B 0, that is: S 0 = (1 + α ) · B 0 . The time constant τ the exponentially decaying disturbance is called the relaxation time constant.

[0004] Depending on the display technology, the disruption caused by burn-in can affect the luminance and / or color of the displayed image. Burn-in is a common occurrence in displays, particularly liquid crystal displays (LCDs) and organic light-emitting diode (OLEDs). Burn-in is also common in displays and projecting display devices based on digital mirror device (DMD) technology. However, other displays and projecting display devices operating according to different physical principles can also be affected by this type of burn-in.

[0005] It is also possible that the disturbance caused by burn-in does not completely disappear, especially after a very long exposure to a burn-in image over many hours or weeks.

[0006] Prior art methods for determining burn-in behavior involve controlling a display device with a test image and measuring the luminance distribution emitted by the display device. The display device is then controlled with a burn-in image for a predetermined period of time, for example, several minutes, hours, or even several days, thereby provoking burn-in. Immediately thereafter, the display device is controlled again with the test image, and the emitted luminance distribution is measured. Due to the effect of burn-in, the resulting luminance distribution differs from the luminance distribution originally measured (before burn-in).

[0007] Devices and methods for both scanning (i.e., point-by-point scanning) and area-wide measurement of a luminance distribution are known. In particular, luminance measuring cameras are known with which the spatial distribution of a photometric parameter, preferably a luminance emitted by a display device, can be measured.

[0008] The difference in the luminance distributions can be measured as a quotient or as a difference. Multiple differences can also be measured by measuring the luminance distribution over time of the test image emitted by the display after the burn-in process.

[0009] For a quantitative determination of both the burn-in effect and the recovery of a display device (i.e. the decay of the disturbance of the display caused by the burn-in), it is necessary to relate the measurement time of a luminance distribution to the time at which the burn-in image was switched off and the test image was switched on, which shall hereinafter be referred to as the start of relaxation.

[0010] Methods are known in which the luminance distribution is measured by recording a camera image from a luminance measuring camera at a time staggered from the activation of the display device with the test image. For example, a computer can first switch the activation of the display device from the burn-in image to the test image via a graphics output, and then, after a predetermined waiting time, the recording of the camera image on the luminance measuring camera can be triggered.

[0011] It is also known that the switching of the image displayed by the display device does not occur synchronously with the switching from the burn-in image to the test image, but rather with a delay. The delay depends, among other things, on the technology of the display device and the type of graphics output used by the computer to control the display device. This delay is generally not precisely known or can only be determined with additional effort. This delay can be variable, so that multiple measurements can produce different values for the delay.

[0012] It is also possible that the switching from the burn-in image to the test image does not occur synchronously for all image pixels, so that some image pixels already display the pixel value assigned to the test image, while other image pixels still display the pixel value assigned to the burn-in image.

[0013] In addition, the time at which a recording can be triggered on a luminance measuring camera is generally not synchronized with the switching of the displayed image, but is subject to a fluctuating deviation (jitter).

[0014] To quantify the burn-in effect, it is advantageous to record at least one luminance distribution immediately after switching the image actually displayed by the display device.

[0015] Due to the unknown effective switching delay and jitter, there is a risk of triggering a camera image on the luminance meter before the displayed burn-in image has switched to the test image. In this case, the luminance meter's exposure is with the burn-in image for at least part of the exposure time. This overestimates the effect of the burn-in.

[0016] On the other hand, there is a risk of the luminance measurement camera being triggered too late relative to the display switching. In this case, the luminance distribution of the display device is recorded in a state that has already partially recovered. This underestimates the effect of burn-in.

[0017] From the publication Rotscholl, Ingo; Krüger, Udo (2019): 50-3: Aspects of Image Sticking Evaluations Using Imaging Luminance Measurement Devices. In: SID Symposium Digest of Technical Papers 50 (1), pp. 695-698. DOI: 10.1002 / sdtp.13014, a method and a device are known in which a photometric parameter is recorded on a display using an external measuring device independent of a luminance measuring camera. If the image displayed on the display varies, for example, when switching from a burn-in image to a test image, the photometric parameter recorded by the external measuring device also varies.By comparing the measured photometric parameter with a predetermined threshold value, determined in particular based on the burn-in image, the test image, the display section captured by the external measuring device, and the integration time for the measurement, the onset of relaxation—that is, the switching from the display of the burn-in image to the display of the test image—can be determined more precisely. This makes it possible to more precisely trigger the recording of a luminance distribution relative to the recovery time or relaxation time of the measured display, or to relate it more precisely mathematically, for example, to determine a relaxation time constant.

[0018] The document Lee, Don-Gyou et al.: "The Analysis Method of Dynamic Characteristics in LCD." In: SID Symposium Digest of Technical Papers, 2003, pp. 292-295, describes a method for analyzing dynamic delay phenomena in liquid crystal displays (LCDs). The method uses a high-speed charged-coupled device (CCD) camera with a resolution of 480 x 420 pixels and a dynamic range of 8 bits, which captures an LCD display with an optical device. To objectify a measure of dynamic delay, a parameter called the dynamic delay contrast (DDC) ratio is proposed.

[0019] Document CN 110 376 218 A describes a method and a device for detecting ghost images on display devices. According to the method, a first image with a predetermined identifying code comprising a first and a second gray value is displayed on the display device, and then switched to a second image with a third gray value lying between the first and second gray values. An image capture device continuously captures the displayed second image. The time up to which the identifying code of the first image can be detected is determined.

[0020] The document US 2003 / 0214586 A1 describes a method and a device for quantitatively determining the image quality of an image displayed by a display device.

[0021] Document US 2007 / 0222860 A1 describes a device for correcting image burn-in on a display. The device comprises an image capture unit for capturing images displayed by the display, and a light control unit for controlling individual pixels of the display to display a predetermined luminance value. The image capture unit captures the luminance difference pixel-by-pixel when displaying the predetermined luminance value. A computing unit compares the detected luminance difference with the predetermined luminance value, and a correction is determined from the resulting difference.

[0022] The invention is based on the object of providing an improved method for determining the onset of relaxation on a pixel-by-pixel controllable optical display device that has been subjected to a burn-in process, which can be implemented particularly simply and with minimal equipment complexity. This object is achieved according to the invention by a method having the features of claim 1.

[0023] Furthermore, the invention is based on the object of providing a method for determining the burn-in behavior of a pixel-by-pixel controllable optical display device. This object is achieved according to the invention by a method having the features of claim 5.

[0024] Furthermore, the invention is based on the object of providing a device for determining the burn-in behavior of a pixel-by-pixel controllable optical display device. This object is achieved according to the invention by a device having the features of claim 10.

[0025] Furthermore, the invention is based on the object of specifying an application of a method for determining the burn-in behavior of a pixel-by-pixel controllable optical display device. This object is achieved according to the invention by an application having the features of claim 12.

[0026] Advantageous embodiments of the invention are the subject of the subclaims.

[0027] In a method for determining a relaxation start of a relaxation which begins after switching an optical display device which can be controlled pixel by pixel from a burn-in image to a relaxation image which is different therefrom, a trigger image area comprising at least one image pixel in the burn-in image is assigned pixel values in such a way that a first parameter formed from the at least one pixel value in the trigger image area of the burn-in image differs from a second parameter formed from the at least one pixel value in the trigger image area of the relaxation image.

[0028] For example, the first and second parameters can be determined as a sum or as an average of a plurality of pixel values of the trigger image area in the burn-in image or in the relaxation image.

[0029] The trigger image area can also comprise multiple image pixels. The image pixels assigned to the trigger image area can then be arranged contiguously. However, it is also possible to combine several non-contiguous trigger image sub-areas in one trigger image area, whose image pixels are then assigned pixel values in such a way that a sum formed across all trigger image sub-areas or an average of pixel values formed across all trigger image sub-areas differs from the sum or the average of the pixel values of the image pixels arranged corresponding to the trigger image sub-areas in the relaxation image.

[0030] The burn-in image and the relaxation image are the same size, meaning they have the same pixel dimensions but differ in pixel values. The trigger image area can be formed, for example, as a rectangular or square sub-area. The trigger image area has the same pixel dimensions in the burn-in image and the relaxation image.

[0031] Using a camera comprising a sensor array of sensor pixels, the spatial distribution of a gray value is continuously recorded over time. The camera is configured to read out all of the sensor pixels at a camera clock rate.

[0032] A trigger subfield of sensor pixels is defined such that the trigger subfield overlaps the trigger image area of a display image shown on the display device.

[0033] Using a trigger clock rate, a trigger characteristic value is continuously determined from the pixel values of the sensor pixels in the trigger subfield. The trigger characteristic value can be determined, for example, as the sum or average of all pixel values of the sensor pixels in the trigger subfield; however, other measures, such as statistical ones, are also possible for determining the trigger characteristic value.

[0034] The course of the gray value detected by the sensor pixels corresponds to the course of the local distribution of a photometric characteristic, for example a luminance, over the display device insofar as a temporal change of such a photometric characteristic in the trigger image area of a display image shown on the display device causes a temporal change of a trigger characteristic value determined from the pixel values of the sensor pixels in the trigger subfield.

[0035] In one embodiment, the camera is designed as a luminance measuring camera and is configured to record the spatial distribution of a photometric parameter.

[0036] A trigger threshold value is determined which lies between the trigger characteristic value resulting from the recording of the burn-in image and the trigger characteristic value resulting from the recording of the relaxation image.

[0037] The start of relaxation is determined as the point in time at which the continuously determined trigger characteristic value crosses the trigger threshold, for example, exceeds it for the first time or falls below it for the first time.

[0038] One advantage of the method is that it operates independently of the latency, which cannot be precisely determined, between the activation of the display device and the switching of the display's image. This allows the relaxation point at which the display device begins to recover from the effects of the burn-in image to be determined more precisely. Another advantage of the method is that it does not require any additional resources beyond the camera. Furthermore, latencies and jitter, which inevitably occur when the camera is controlled by such additional resources, can be avoided.

[0039] In one embodiment of the method, the trigger clock rate is selected to be higher than the camera clock rate intended for reading the sensor pixels of the camera's sensor field. An advantage of this embodiment is that the relaxation time can be determined more precisely than is possible by evaluating the individual images provided by the camera.

[0040] It is possible to increase the trigger clock rate compared to the camera clock rate by only reading out partial areas of the sensor field in a form that is sufficient and suitable for determining the trigger characteristic. For example, it is possible to combine a plurality of neighboring sensor pixels using sensor pixel binning and read them out as a single value that corresponds to the sum or average of all the combined sensor pixels. Methods in which the camera's sensor field is undersampled, for example by only reading out every second or every fourth, or generally every nth, sensor pixel, are also possible. Methods in which only one partial area of the camera's sensor field, known as the area of interest (AOI), or several such partial areas are read out are also possible and lead to a significant increase in the trigger clock rate.

[0041] In one embodiment of the method, only a subfield of the camera's sensor field is read out, which includes at least the trigger subfield. The sensor pixels assigned to the trigger subfield and thus also to the read subfield of the sensor field can be arranged contiguously. However, these sensor pixels can also be distributed non-contiguously across the sensor field. The read subfield of the sensor field can be limited to the trigger subfield. However, it can also include further, additional sensor pixels that are not assigned to the trigger subfield.

[0042] In this embodiment, the trigger clock rate is selected to be higher than the camera clock rate intended for the complete readout of the sensor field.

[0043] Such methods ensure that a smaller number of values need to be transmitted when reading a portion of the sensor field that includes the trigger subfield, allowing the trigger clock rate to be increased relative to the camera clock rate. This allows the relaxation time to be determined even more precisely.

[0044] For example, using such methods, a trigger clock rate of over 1 kilohertz and thus a temporal resolution in the millisecond range when determining the relaxation time can be achieved at a camera clock rate of 20 hertz.

[0045] In one embodiment of the method, the trigger characteristic value is only recorded in a time range of predetermined duration, which includes the time at which the control of the display device is switched from the burn-in image to the relaxation image. The switching time at which the control of the display device is switched from the burn-in image to the relaxation image can be determined only imprecisely, but with little technical effort. To achieve a high degree of accuracy regarding the onset of relaxation, it is then sufficient to examine a certain time range before and / or after this switching time in more detail. This can reduce the effort required to implement the method.

[0046] In a further embodiment of the method, an average trigger characteristic is formed from a plurality of trigger characteristic values that are successively acquired in a time range of predetermined duration. This average trigger characteristic is compared with the trigger threshold instead of the individual trigger characteristic values. The average trigger characteristic can be formed using a moving average, a moving median, a low-pass filter, or a similar signal processing method for eliminating high-frequency interference.

[0047] This embodiment has the advantage of eliminating fluctuations in the continuously determined trigger characteristic value, which may be caused by the brightness modulation of the display device. For example, display devices can control the brightness of image pixels using pulse-width modulation and thus generate varying trigger characteristic values, depending on the temporal position of the camera's integration time relative to the pulse-width modulation clock, for the same image content displayed on the display device. The proposed formation of the average trigger characteristic value prevents erroneous detection of sweeping, i.e., exceeding or falling below the trigger threshold.

[0048] In one embodiment of the method, the pixel values in the trigger image area are selected such that a first mean value of pixel values in the trigger image area of the burn-in image is higher than a second mean value of pixel values in the trigger image area of the relaxation image. Preferably, the pixel values in the trigger image area of the burn-in image are selected such that the maximum achievable luminance with the display device is achieved in at least a partial area of the trigger image area.

[0049] This allows for a reduction in the integration or exposure time during which the trigger image area must be exposed to achieve an evaluable trigger characteristic. This allows for a more dense temporal sequence (or sampling) of the trigger characteristic. This allows for more precise determination of the onset of relaxation.

[0050] In a method for determining the burn-in behavior of a pixel-by-pixel controllable optical display device, the display device is controlled to display a reference image if it has not been subjected to a burn-in process. A camera captures a first camera image that captures the spatial distribution of a gray value corresponding to the spatial distribution of a photometric parameter across the display device, while the display device is controlled to display the reference image.

[0051] Subsequently, the display device is controlled for a predetermined burn-in time to display a burn-in image.

[0052] The display device is then controlled to display a relaxation image. At least one additional camera image is captured by the camera, which captures the spatial distribution of a gray value corresponding to the spatial distribution of a photometric parameter across the display device while the display device is controlled to display the relaxation image.

[0053] According to the invention, a relaxation start is determined by means of the camera according to one of the methods described above and the burn-in behavior for the display device is determined from the first camera image, the at least one further camera image and the relaxation start.

[0054] An advantage of this method is that the onset of relaxation can be determined more precisely, and thus the relaxation behavior, for example, the relaxation time constant, of the display device, which results from the disturbance in the measured image related to the relaxation duration, can also be determined more precisely. A further advantage of the method is that it does not require any additional resources beyond the camera. This also avoids latencies and jitter that inevitably occur when controlling the camera using such additional resources. This enables a more reliable and accurate assessment of the relaxation behavior of the display device.

[0055] In one embodiment of a method for determining burn-in behavior, only a subfield of the sensor field, i.e., only a portion of all the camera's sensor pixels, is read from the at least one additional camera image, which captures the spatial distribution of a gray value corresponding to the spatial distribution of a photometric parameter across the display device while the display device is controlled to display the relaxation image. This subfield is read at a readout clock rate that is higher than the camera clock rate provided for the complete readout of the sensor field (i.e., for the readout of all the camera's sensor pixels).

[0056] This allows the progression of the photometric parameter during the relaxation image display to be recorded with higher temporal resolution than with a complete readout of at least one additional camera image. This can improve the accuracy of the burn-in evaluation.

[0057] In one embodiment, the subfield that is read out in the at least one additional camera image during the display of the relaxation image on the display device at the readout clock rate that is higher than the camera clock rate is selected to at least partially overlap the trigger subfield used to determine the trigger characteristic value. This simplifies the implementation of the method and reduces the effort required to establish the relationship between the subfield and the corresponding image area of the relaxation image.

[0058] In one embodiment of the method, the camera is designed as a luminance measuring camera, with which a particularly accurate, in particular quantitative, determination of the burn-in behavior of a display device is possible.

[0059] In one embodiment of the method, the at least one further camera image is recorded at a predetermined time offset from the point in time at which the trigger characteristic value exceeds the trigger threshold value.

[0060] For some types of display devices, it is known that the switching from a first image, for example, the burn-in image, to a subsequent second image, for example, the relaxation image, does not occur instantaneously at the time the new image content is transmitted to the display device. Rather, in such display devices, the switching occurs for different image pixels at different times, distributed within a switching time interval that is typically a few milliseconds to 100 milliseconds long. The switching time interval is known or can be determined depending on the type of display device.

[0061] In the present embodiment of the method, to determine the onset of relaxation, a time offset is added at the time at which the trigger characteristic value crosses the trigger threshold, which at least covers the known or previously determined switching time interval. This ensures that at the thus determined onset of relaxation, all image pixels (and not just those of the trigger image area) display the image content of the relaxation image.

[0062] An advantage of this embodiment is that the effect of burn-in can be more easily compared and evaluated if camera images are recorded at a predetermined time interval from the onset of relaxation and used for evaluation. In particular, this prevents the effect of burn-in from being incorrectly overestimated due to an image displayed by the display device at a relaxation onset determined too early still representing the burn-in image in parts, particularly outside the trigger image area.

[0063] A device for determining the burn-in behavior of a pixel-by-pixel controllable optical display device comprises a control unit and a camera. The camera is designed as a complementary metal-oxide semiconductor (CMOS) camera. Alternatively, the camera can also be designed as a charge-coupled device (CCD) camera. The control unit is configured to control a pixel-by-pixel controllable optical display device. According to the invention, the control unit and the camera are configured to carry out the method described above for determining the burn-in behavior of a pixel-by-pixel controllable optical display device.

[0064] The camera can be designed as an intelligent camera (smart camera) and comprise an internal processor which is configured to carry out image processing operations and / or to carry out the method described above for determining the burn-in behavior of a pixel-by-pixel controllable optical display device.

[0065] One advantage of this device is that the onset of relaxation and thus the burn-in behavior can be determined particularly accurately and reproducibly. A further advantage is that the technical complexity and the effort required to set up and calibrate the measuring device can be reduced compared to state-of-the-art measuring devices, which use an external measuring device independent of the camera to determine the onset of relaxation.

[0066] In one embodiment of the device, the camera is configured to continuously determine the trigger characteristic value from the pixel values of the sensor pixels in the trigger subfield and to compare the trigger characteristic value with the trigger threshold value.

[0067] An advantage of this embodiment is that the determination of the onset of relaxation and the associated triggering of one or more camera images can be performed independently within the camera. This avoids or reduces the influence of latency and jitter, which inevitably occur during data exchange between the camera and the computer. This improves the accuracy and reproducibility of determining burn-in behavior. Furthermore, the equipment complexity can be reduced, as no external measuring devices are required.

[0068] In a further aspect of the invention, the proposed method for determining the burn-in behavior of a pixel-by-pixel controllable optical display device is used to determine the burn-in behavior of a display intended for use in a vehicle.

[0069] Displays used in vehicles are used for a particularly long period of operation and, to a particularly high extent, for displaying static or long-term unchanged graphics or images. Therefore, the precise and reliable determination of disturbances caused by such static images is particularly important. The proposed method enables a more reliable evaluation of displays than prior art methods due to a more precise determination of the onset of relaxation.

[0070] Embodiments of the invention are explained in more detail below with reference to the drawings, in which: Figure 1 schematically shows a device for measuring the burn-in behavior of a display, Figure 2 schematically shows a sequence of images projected during the measurement of the burn-in behavior, Figure 3 schematically shows a device for measuring the burn-in behavior of a display, Figure 4 schematically shows a sensor field of a CMOS camera and Figure 5 schematically shows the course of a trigger integral value corresponding to a trigger image area.

[0071] Corresponding parts are provided with the same reference numerals in all figures.

[0072] Figur 1 shows schematically a measuring setup known from the prior art for measuring the burn-in behavior of a display 1. The measuring setup comprises a computer 2 which is connected to the display 1 and to a camera 3.

[0073] The display 1 to be measured can be designed, for example, as a liquid crystal display (LCD liquid crystal display) or as an organic light emitting diode (OLED).

[0074] Computer 2 is configured to display images or graphics on display 1. Instead of computer 2, a test image generator (not shown in detail) can also be used to display images on display 1.

[0075] This image shows the output of a burn-in image EB on display 1. The burn-in image EB is designed as a checkerboard pattern with square fields distributed across the entire image area. The square fields each have an approximately homogeneous luminance, with adjacent fields differing in luminance. The checkerboard pattern is preferably composed of fields with the maximum contrast difference.

[0076] The connection between the computer 2 or the test image generator (not shown in detail) and the display 1 can be implemented, for example, as a VGA (Video Graphics Array) connection, an HDMI (High Definition Multimedia Interface) connection, a DisplayPort connection, or a DVI (Digital Visual Interface) connection. Generic bus types, whose use is not limited to the transmission of images or graphics, can also be used to connect the computer 2 or a test image generator to the display 1. For example, a Controller Area Network (CAN) bus can be used for the connection.

[0077] Camera 3 is designed to record a Figur 1 The camera image KB (not shown in detail) is configured to capture the burn-in image EB displayed by the display 1. The camera 3 is configured to capture the luminance emitted by the display 1 pixel by pixel as a gray value or as a color value. In particular, the optical axis of the camera 3 is aligned perpendicular to the display 1. Furthermore, the camera 3 is configured to provide photometric or colorimetric pixel values for a plurality of image pixels.

[0078] The computer 2 is configured to receive and evaluate a camera image KB taken by the camera 3.

[0079] Figur 2 shows a schematic of the sequence of images that are displayed on display 1 according to the state of the art when measuring burn-in behavior. The sequence is divided into three phases P1, P2, and P3.

[0080] In a first phase P1, the display 1 is controlled by the computer 2 with a rolling sequence of warm-up images WB1 to WB3. The warm-up images WB1 to WB3 have an approximately homogeneous grayscale distribution with different grayscale values. The warm-up images WB1 to WB3 are designed so that the display 1 reaches a steady operating state and that no burn-in effects are triggered on the display 1. The sequence of warm-up images WB1 to WB3 is in Figur 2 only shown once, but can be repeated cyclically if necessary to achieve a steady-state operating state.

[0081] At the end of the first phase P1, display 1 is controlled with a reference image B0 which has a homogeneous reference gray value.

[0082] In a second phase P2, which follows the first phase P1, the display 1 is controlled with the burn-in image EB, which is already based on Figur 1 was described.

[0083] In a third phase P3, which follows the second phase P2, the display 1 is controlled with a relaxation image RB. Preferably, the relaxation image RB is identical to the reference image B0.

[0084] For each display image DB shown by the display 1, at least one camera image KB(DB) is recorded by the camera 3. A burn-in effect, in particular a relaxation time constant τ can be determined from the comparison of the camera image KB(B0) assigned to the reference image B0 with the camera image KB(RB) assigned to the relaxation image RB.

[0085] For an accurate determination of the burn-in effect, it is advantageous if the time interval between the display of the relaxation image RB on the display 1 and the recording of the associated camera image KB(RB) is known as accurately as possible.

[0086] The control of the display 1 by the computer 2, i.e. the output of a different image via the connection between the computer 2 and the display 1, takes place at control times t 1 , t 3 , ... t 9 . Due to the signal transmission from computer 2 to display 1, due to the inertia of display 1 and due to the exposure time of camera 3, which is usually a few hundredths to tenths of a second, and the clock for reading the camera images KB, which is offset from the clock of the display images DB, the display images DB provided by computer 2 appear in camera 3 with a delay at switching times t 2 , t 4 , . . . t 10 as camera images KB.

[0087] Using methods known from the state of the art, the latency Δ t i , i + 1 = t i + 1 − t i , i = 1 , 3 , 5 , 7 , 9 Therefore, the time between the switching of the display image DB by computer 2 and the time of recording the camera image KB assigned to the new display image DB cannot be precisely determined. This also affects the accuracy of determining the burn-in effect.

[0088] The present invention has recognized and overcome this disadvantage, as can be seen from the Figur 3 depicted, compared to Figur 1 modified measurement setup is explained below.

[0089] The measurement setup includes a modified camera 13 and a modified computer 12. The computer 12 is configured to display a modified burn-in image EB' on the display 1. The modified burn-in image EB' has a trigger image area TB with a homogeneous grayscale distribution. Preferably, the trigger image area TB covers a plurality of fields of the checkerboard pattern and has a grayscale that differs as much as possible from the average grayscale of the burn-in image EB' as well as from the average grayscale of the relaxation image RB. For example, the trigger image area TB can have the maximum (brightest) grayscale displayable on the display 1.

[0090] The modified camera 13 is configured to read and / or evaluate a predefined sub-area of a camera image KB. In particular, the modified camera 13 is configured to read this sub-area at a higher speed and at a higher frequency than the complete camera image KB. Changes in the camera image KB captured by the predefined sub-area can therefore be detected particularly quickly.

[0091] In a Figur 4 In the illustrated embodiment, the camera 13 is designed as a CMOS (complementary metal-oxide-semiconductor) camera and has a sensor field 14 with a plurality of matrix-like arranged light-sensitive sensor pixels 15. The digitized measured values of the sensor field 14 form the camera image KB.

[0092] A column address decoder 16 and a row address decoder 17 select a sensor pixel 15 for output whose address matches the address specified by an address generator 18. The selected sensor pixel 15 is provided as a digital value in a readout register 19. Similarly, a range of sensor pixels 15 of a row of the sensor array 14 can be provided as a plurality of digital values in the readout register 19.

[0093] This makes it possible to read out all sensor pixels 15 located in a square or rectangular trigger subfield 20 of the sensor field 14 particularly quickly, in particular much faster and at much shorter time intervals than the entirety of the sensor pixels 15 of the sensor field 14.

[0094] The camera 13 further comprises a camera controller 21 which is connected to the address generator 18 and to the readout register 19 and which has an interface for exchanging data with the computer 2.

[0095] In the following, the inventive operation of the Figur 3 described camera 13.

[0096] A camera image KB is transmitted to the computer 2 as the totality of all digitized pixel values of the sensor pixels 15 of the sensor field 14. In the computer 2, a partial area of pixels of the camera image KB is identified that covers the trigger image area TB of the burn-in image EB' shown on the display 1. Preferably, a rectangular or square partial area of the camera image KB is determined that covers the largest possible area of the trigger image area TB.

[0097] This partial area can be determined by automatic image processing. The partial area in the camera image KB can also be determined by manual marking by displaying the camera image KB on a display device (not shown in detail) and selecting a rectangle or square inscribed in the trigger image area TB, for example, using a pointing device.

[0098] From the specification of the sub-area in the camera image KB, the address range of the trigger sub-area 20 can be determined by transferring the coordinates of pixels of the camera image KB to row addresses and column addresses of sensor pixels 15 of the sensor field 14 and transmitted to the camera controller 21. For example, the indices of the first and last column and the first and last row of the sensor field 14, which delimit the trigger sub-area 20, can be transmitted.

[0099] The camera controller 21 is configured and programmed such that the digitized pixel values of the trigger subfield 20 are continuously read out, and a trigger characteristic value T is determined from the pixel values of the trigger subfield 20. For example, the sum or mean of the pixel values of the trigger subfield 20 can be determined as the trigger characteristic value T.

[0100] Due to the very small trigger subfield 20 compared to the sensor field 14, the trigger characteristic value T can be determined at very short time intervals.

[0101] Figur 5 shows the course of the trigger characteristic value T determined in this way over the time axis t. Since the trigger subfield 20 is matched to the trigger image area TB of the display image DB shown on the display 1, the time course of the trigger characteristic value T has particularly high values when the burn-in image EB' with the superimposed trigger image area TB is shown on the display 1 with a particularly high (light) gray value.

[0102] This makes it possible to determine the switching time t 10 , at which the display 1 switches between the display of the burn-in image EB' and the display of the relaxation image RB. For example, a trigger integral threshold value can be determined based on the particularly high (bright) gray value displayed in the trigger image area TB and on the extent of the trigger image area TB as well as on the exposure time during which the trigger image area TB is exposed. T S If the trigger characteristic value T determined by the camera control 21 exceeds the trigger integral threshold T S , this excess indicates that the burn-in image EB' was recorded by the camera 13.

[0103] Accordingly, the following first undershoot of the trigger integral threshold shows T S at the time t R that the relaxation image RB is shown on the display 1 and is recorded by the camera 13. This time can be considered as the start of relaxation t R at which the relaxation of the display 1 begins and to which the temporally decreasing deviation of the camera image KB(RB) recorded during the third phase P3 (relaxation phase) compared to the camera image KB(B0) recorded at the end of the first phase P1 (warm-up phase) is to be related.

[0104] The recording of the camera image KB(RB), which is a t R The relaxation image RB shown on display 1 is recorded and triggered by a trigger signal. The trigger signal can be used directly at the detected start of relaxation t R Preferably, the trigger signal is generated after a predetermined delay or waiting time Δ t R , that is: at the time t R + Δ t R , generated, whereby this delay, as explained in more detail below, depends on the switching time interval [ t 9 , t 10 ] is selected, which is known or determined for the display device.

[0105] An advantage of the method according to the invention is that the accuracy with which the start of the relaxation of the display 1 can be determined is essentially only limited by the accuracy with which the trigger integral threshold value T S is determined, as well as from the time interval in which all sensor pixels 15 within the trigger subfield 20 can be read out and summed up by the camera 13.

[0106] Compared to the state of the art, latency and jitter play no role in the transmission of a display image DB from computer 2 to display 1. The inertia of display 1 also influences the accuracy in determining the onset of relaxation. t R not or only insignificantly.

[0107] By more precisely determining the onset of relaxation t R It is also possible to determine the burn-in effect more accurately than with state-of-the-art methods. For example, a relaxation time constant τ be determined more reliably and accurately.

[0108] In one embodiment, the determination of the trigger characteristic value T can only be carried out in a time interval around the activation time t9 can be executed when the relaxation image RB is transmitted from the computer 2 to the display 1 after the burn-in image EB'. This time interval can be limited roughly, for example, to a few tenths of a second. For example, the computer 2 can be programmed so that immediately before the display image DB changes from the burn-in image EB' to the relaxation image RB, a signal is transmitted to the camera control 21, which triggers the determination of the trigger characteristic value T for a predetermined period of time, for example, 500 milliseconds.

[0109] The camera control 21 can also be configured so that the exposure of the camera image KB(RB) related to the relaxation image RB after a predetermined waiting time Δ t R from the beginning of relaxation t R begins. The predetermined waiting time Δ t R can, for example, be entered by the user on the computer 2 and transmitted from the computer 2 to the camera control 21.

[0110] It is possible to set the predetermined waiting time Δ t R to be set so that the switching process of the display 1 from the display of the burn-in image EB' to the display of the relaxation image RB is reliably completed when the camera image KB(RB) related to the relaxation image RB is exposed. Preferably, the waiting time Δ t R set equal to or slightly larger than the switching time interval [ t 9 , t 10 ] . The [ t 9 , t 10 ] indicates the time range in which the image pixels of the displayed image are reliably switched from the burn-in image EB to the relaxation image RB.

[0111] This prevents a camera image KB from being recorded at a time when the burn-in image EB is still partially displayed by the display device 1, particularly in image areas outside the trigger image area TB, and thus overestimating the effect of the burn-in. This improves the reliability of the evaluation of the relaxation behavior of the display 1.

[0112] It is also possible to set the predetermined waiting time Δ t R according to criteria that cover specific requirements when using the display 1 and / or that may be specified by test standards.

[0113] In one embodiment, the burn-in behavior of the display 1, for example the relaxation time constant τ , can be determined based on a partial area of the camera image KB(RB) recorded corresponding to the relaxation image RB. For example, the burn-in behavior can be determined by evaluating only those image pixels of the camera image KB(RB) that are provided by the sensor pixels 15 within the trigger subfield 20, wherein the burn-in image EB' and the relaxation image RB are selected such that the sum or the mean value of the image pixels respectively assigned to the trigger image area TB differs between the burn-in image EB' and the relaxation image RB.

[0114] However, it is also possible that, in order to determine the burn-in behavior, one sensor pixel 15 or several sensor pixels 15 of the camera image KB(RB) recorded from the relaxation image are read out and compared with the camera image KB(EB) of the burn-in image EB, which are not or only partially located in the trigger subfield 20.

[0115] For example, the burn-in image EB' can have a checkerboard pattern with dark (black) and light (white) fields, with one or more light (white) fields of the checkerboard pattern being assigned to the trigger image area TB. The reference image B0 is selected to be homogeneous with a medium (gray) gray value.

[0116] The burn-in behavior is then determined by evaluating the sensor pixels 15 arranged in the trigger subfield 20 or another subfield of the sensor field 14, onto which the selected white field(s) are imaged during the burn-in (second phase P2), and then during the relaxation (third phase P3), the average gray values disturbed by the burn-in. It is also possible to assign selected dark (black) fields of the checkerboard pattern to the trigger image area TB instead of the selected light (white) fields of the checkerboard pattern.

[0117] An advantage of this embodiment is that by reducing the number of pixels, the determination of whether the trigger characteristic value T determined over the trigger image area TB exceeds the trigger threshold T S This allows for a higher temporal resolution in determining the point in time at which the trigger characteristic value T exceeds the trigger threshold T S This can be the beginning of relaxation t R determined more precisely and the burn-in behavior of the display 1 can be determined more precisely.

[0118] Furthermore, by reading only a partial field of the sensor field 14 during relaxation (third phase P3), a readout clock rate is achieved that is higher than the camera clock rate intended for the complete readout of all sensor pixels 15 of the camera 13. This allows the temporal progression of the gray values displayed by the display 1, potentially disturbed by burn-in, to be recorded with a higher temporal resolution during relaxation, and from this, the burn-in behavior of the display 1 can be determined more precisely.

[0119] It is possible, but not necessary, that the subfield of the sensor field 14 read out during the relaxation is selected to be identical or overlapping with the trigger subfield 20, which is used for the continuous determination of the trigger characteristic value T and for determining the start of relaxation t R is evaluated. It is also possible, but not required, for the subfield read during relaxation to be contiguous. The subfield is selected so that it can be read at a higher readout rate than the camera clock rate. This results in the advantage of higher temporal resolution when determining the relaxation behavior, and thus when evaluating the burn-in behavior for display 1.

[0120] In an alternative embodiment, the evaluation of the image pixels of the camera image KB can also be performed on the computer 2, wherein the camera controller 21 is configured such that the camera image KB or the part of the camera image KB corresponding to the trigger image area TB is transmitted to the computer 2. This embodiment has the advantage that the evaluation of the camera image KB can be implemented independently of the specifically used camera 13 and that thus different types of cameras 13 can be used flexibly.

[0121] In contrast, the embodiment in which the sensor values of the sensor field 14 and / or the trigger subfield 20 are determined by the camera control 21 has the advantage that less data must be transmitted between the camera 13 and the computer 2. This results in lower latency and less jitter between the control time t9 the switch to the relaxation image RB and the determined relaxation start t R possible. BEZUGSZEICHENLISTE

[0122] 1Display, indicator 2, 12Computer, control unit 3, 13Camera 14Sensor field 15Sensor pixel 16Column address decoder 17Row address decoder 18Address generator 19Readout register 20Trigger subfield 21Camera control B0Reference image P1First phase, warm-up phase P2Second phase, burn-in phase P3Third phase, DBDisplay image, display image EB, EB'Burn-in image KBCamera image RBRelaxation image TBTrigger image range TTrigger characteristic value tTime axis t 1 , t 3 , t 5 , t 7 , t 9 Control time t 2 , t 4 , t 6 , t 8 , t 10 Switchover time t R Relaxation onset, time Δ t R Waiting time WB1, WB2, WB3 first to third warm-up image

Claims

1. Method for determining a start of relaxation (tR) when switching a pixel-by-pixel controllable optical display device (1) from a burn-in image (EB') to a relaxation image (RB), wherein - in the burn-in image (EB') a trigger image area (TB) comprising at least one image pixel is allocated pixel values such that a first parameter formed over the at least one pixel value in the trigger image area (TB) of the burn-in image (EB') differs from a second parameter formed in the same way over the at least one pixel value in the trigger image area (TB) of the relaxation image (RB), - by means of a camera (3, 13) comprising a sensor field (14) of sensor pixels (15), the spatial distribution of a greyscale value corresponding to the spatial distribution of a photometric characteristic variable over the display device (1) is recorded continuously over time, wherein the camera (3, 13) is configured for reading out the totality of all sensor pixels (15) with a camera clock rate and subareas of the sensor field (14) can be read out with a higher clock rate, wherein o a trigger subfield (20) comprising at least one sensor pixel (15) is defined overlapping with respect to the trigger image area (TB) of a display image (DB) presented on the display device (1), and o with a trigger clock rate, a trigger characteristic value (T) is determined continuously over time from the pixel values of the at least one sensor pixel (15) in the trigger subfield (20), o wherein the trigger subfield (20) is defined as a subarea of the sensor field (15) and is read out with the trigger clock rate, and - a trigger threshold value (TS) is determined, which lies between the trigger characteristic value (T) resulting for the recording of the burn-in image (EB') and the trigger characteristic value (T) resulting for the recording of the relaxation image (RB), and - the start of relaxation (tR) is ascertained as the point in time at which the continuously ascertained trigger characteristic value (T) crosses the trigger threshold value (TS), wherein the trigger clock rate is chosen to be higher than the camera clock rate provided for reading out the totality of the sensor pixels (15) of the sensor field (14).

2. Method according to Claim 1, wherein the trigger characteristic value (T) is only recorded in a time period of predetermined duration, which comprises the point in time at which the control of the display device (1) is switched from the burn-in image (EB') to the relaxation image (RB).

3. Method according to Claim 2, wherein a moving average value is determined from a plurality of successively ascertained trigger characteristic values (T) and is compared with the trigger threshold value (Ts).

4. Method according to any of the preceding claims, wherein the pixel values in the trigger image area (TB) are chosen such that a first average value of pixel values in the trigger image area (TB) of the burn-in image (EB') is higher than a second average value of pixel values in the trigger image area (TB) of the relaxation image (RB), and are preferably chosen such that the maximum luminance achievable with the display device (1) is achieved at least in a subarea of the trigger image area (TB) of the burn-in image (EB').

5. Method for determining the burn-in behaviour of a pixel-by-pixel controllable optical display device (1), wherein - the display device (1), if it has not been subjected to a burn-in process, is controlled for presenting a reference image (B0), wherein - by means of a camera (3, 13), a first camera image (KB(B0)) is recorded, which records the spatial distribution of a greyscale value corresponding to the spatial distribution of a photometric characteristic variable over the display device (1) while the latter is controlled for presenting the reference image (B0), - the display device (1) is controlled for presenting a burn-in image (EB') for a predetermined burn-in period, - the display device (1) is controlled for presenting a relaxation image (RB), wherein o by means of the camera (3, 13), at least one further camera image (KB(RB)) is recorded, which records the spatial distribution of a greyscale value corresponding to the spatial distribution of a photometric characteristic variable over the display device (1) while the latter is controlled for presenting the relaxation image (RB), and o a start of relaxation (tR) is ascertained by means of the camera (3, 13) according to a method of the preceding claims, and - the burn-in behaviour for the display device (1) is ascertained from the first camera image (KB(B0)), the at least one further camera image (KB(RB)) and the start of relaxation (tR).

6. Method according to Claim 5, wherein from the at least one further camera image (KB(RB)), which records the spatial distribution of a greyscale value corresponding to the spatial distribution of a photometric characteristic variable over the display device (1) while the latter is controlled for presenting the relaxation image (RB), a subfield of the sensor field (14) is read out with a readout clock rate which is higher than the camera clock rate provided for the complete readout of the sensor field (14).

7. Method according to Claim 6, wherein the subfield read out with the readout clock rate at least partially overlaps the trigger subfield (20).

8. Method according to any of Claims 5 to 7, wherein the camera (3, 13) is designed as a luminance measuring camera (13).

9. Method according to any of Claims 5 to 8, wherein the at least one further camera image (KB(RB)) is recorded within a predetermined temporal offset with respect to the ascertained start of relaxation (tR).

10. Device comprising a control unit (2, 12) and a camera (3, 13), wherein the control unit (2, 12) is configured for controlling a pixel-by-pixel controllable optical display device (1), wherein the control unit (2, 12) and the camera (3, 13) are configured for carrying out a method according to any of Claims 4 to 6 and wherein the camera (3, 13) is designed as a complementary metal oxide semiconductor (CMOS) camera or as a charged coupled device (CCD) camera.

11. Device according to Claim 10, wherein the camera (3, 13) is configured for continuously determining the trigger characteristic value (T) from the at least one pixel value of the sensor pixels (15) in the trigger subfield (20) and for comparing the trigger characteristic value (T) with the trigger threshold value (TS).

12. Application of a method according to any of Claims 5 to 9 for determining the burn-in behaviour of a display (1) provided for use in a vehicle.