LED display screen brightness uniformity correction method and system based on current feedback

Abstract

The invention discloses an LED display screen brightness uniformity correction method and system based on current feedback, and relates to the technical field of brightness correction.The method comprises the steps that firstly, a vertical blanking period time sequence empty window is utilized, high-low double-amplitude detection current is injected into LED pixels, and voltage responses in different excitation states are obtained on the premise that normal display is not interfered; then, constructing a state decoupling model based on a reference I-V-T characteristic spectrum, and deeply analyzing coupled electrical signals into independent real-time junction temperature and aging factors by using differential sensitivity characteristics of voltage to temperature and aging; and then, on the basis of the physical state parameters, reverse compensation operation is performed in combination with an original gray instruction and a reference L-I brightness model, and target driving current capable of counteracting thermal quenching and aging attenuation is dynamically deduced. In this way, the correction problem caused by deep coupling of temperature drift and device aging can be effectively solved, and therefore self-adaptive high-precision brightness uniformity correction in the full life cycle of the display screen is achieved.

Description

Technical Field

[0001] This application relates to the field of brightness correction technology, and more specifically, to a method and system for correcting the brightness uniformity of an LED display screen based on current feedback. Background Technology

[0002] With the widespread application of light-emitting diode (LED) display technology in ultra-high-definition video walls, outdoor commercial displays, and immersive virtual studios, the market has set extremely stringent standards for color reproduction and full-screen brightness uniformity. However, due to the discreteness of semiconductor processes and differences in driving circuits, LED pixels exhibit inconsistencies from the outset. Although this can be initially mitigated through point-by-point calibration at the factory, LEDs exhibit significant nonlinear time-varying characteristics over their long service life, with their brightness output depth depending on real-time junction temperature fluctuations and cumulative aging. Traditional static calibration data cannot adapt to this dynamic evolution, inevitably leading to uniformity degradation of the display over time. Given the limitations of external optical camera acquisition solutions, such as susceptibility to ambient light interference, the need to interrupt display, and difficulty in real-time closed-loop control, constructing a dynamic compensation mechanism based on internal electrical parameters (especially current feedback) has become a key path to maintain display quality throughout its entire lifecycle.

[0003] In existing technologies, some driving solutions attempt to achieve brightness compensation by introducing current feedback or combining external thermistor monitoring, but they still face severe challenges in high-precision calibration. The core technical pain point is that the brightness decay of LEDs is a combined result of two physical processes: "reversible temperature drift" and "irreversible material aging," and these two processes are strongly coupled in their electrical characteristics. Specifically, an increase in junction temperature leads to intensified thermal motion of charge carriers, reducing the forward voltage drop and causing a decrease in light conversion efficiency; similarly, device aging leads to a decrease in luminous efficiency and drift in electrical parameters. Most existing feedback schemes are based on simple linear superposition or independent mapping, lacking a unified physical model to deeply analyze these two concurrent factors. This method cannot accurately separate the proportion of voltage or brightness changes attributable to instantaneous thermal effects and the proportion attributable to long-term aging. If the two state variables of "junction temperature" and "aging factor" cannot be effectively decoupled, the compensation algorithm is prone to misjudgment. For example, it may misjudge brightness decrease caused by poor heat dissipation as aging, thus excessively increasing the driving current, which accelerates device decay and fails to meet the precise control requirements of high-end displays.

[0004] Therefore, there is an urgent need for an optimized method and system for correcting the brightness uniformity of LED displays based on current feedback. Summary of the Invention

[0005] This application is made in order to solve the above-mentioned technical problems.

[0006] According to one aspect of this application, a method for correcting the brightness uniformity of an LED display screen based on current feedback is provided, comprising:

[0007] In response to the detection of the vertical blanking period synchronization signal, a low-amplitude probe current and a high-amplitude probe current are injected into the target LED pixel to obtain the measured voltage under the low-amplitude probe current and the measured voltage under the high-amplitude probe current.

[0008] Based on the baseline IVT characteristic spectrum, state decoupling and real-time parameter estimation are performed on the measured voltage under low amplitude probe current, measured voltage under high amplitude probe current, low amplitude probe current, and high amplitude probe current to obtain the estimated real-time junction temperature and estimated aging factor.

[0009] Extract the raw grayscale values ​​of the target LED pixels from the video data stream;

[0010] Based on the baseline LI brightness model, the estimated real-time junction temperature, the estimated aging factor, and the original gray value are used to dynamically calculate the target current to obtain the dynamically adjusted target current.

[0011] Closed-loop drive and PWM correction are performed on the original grayscale value and the dynamically adjusted target current to obtain the corrected PWM duty cycle.

[0012] According to another aspect of this application, a current feedback-based LED display brightness uniformity correction system is provided, comprising:

[0013] The measured voltage acquisition module is used to inject low-amplitude probe current and high-amplitude probe current into the target LED pixel in response to the detection of the vertical blanking period synchronization signal to obtain the measured voltage under the low-amplitude probe current and the measured voltage under the high-amplitude probe current.

[0014] The state decoupling and parameter estimation module is used to perform state decoupling and real-time parameter estimation on the measured voltage under low amplitude probe current, measured voltage under high amplitude probe current, low amplitude probe current, and high amplitude probe current based on the benchmark IVT characteristic spectrum, so as to obtain the estimated real-time junction temperature and the estimated aging factor.

[0015] The raw grayscale value extraction module is used to extract the raw grayscale values ​​of the target LED pixels from the video data stream;

[0016] The dynamic target current calculation module is used to calculate the target current dynamically based on the estimated real-time junction temperature, the estimated aging factor, and the original gray value using the reference LI brightness model, so as to obtain the dynamically adjusted target current.

[0017] The closed-loop drive and PWM correction module is used to perform closed-loop drive and PWM correction on the original grayscale value and the dynamically adjusted target current to obtain the corrected PWM duty cycle.

[0018] Compared with existing technologies, this application provides a current feedback-based method and system for LED display brightness uniformity correction. First, it utilizes a vertical blanking period timing window to inject high and low amplitude probe currents into LED pixels, acquiring voltage responses under different excitation states without interfering with normal display. Then, it constructs a state decoupling model based on a reference IVT characteristic spectrum, leveraging the differential sensitivity of voltage to temperature and aging to deeply analyze the coupled electrical signals into independent real-time junction temperature and aging factors. Furthermore, based on these physical state parameters, it performs inverse compensation calculations using the original grayscale command and a reference LI brightness model, dynamically deriving the target driving current capable of offsetting thermal quenching and aging degradation. This effectively solves the correction problem caused by the deep coupling of temperature drift and device aging, thereby achieving adaptive high-precision brightness uniformity correction throughout the entire lifecycle of the display. Attached Figure Description

[0019] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0020] Figure 1 This is a flowchart of an LED display brightness uniformity correction method based on current feedback according to an embodiment of this application.

[0021] Figure 2 This is a data flow diagram of an LED display brightness uniformity correction method based on current feedback according to an embodiment of this application.

[0022] Figure 3 This is a flowchart of sub-step S1 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application.

[0023] Figure 4 This is a flowchart of sub-step S2 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application.

[0024] Figure 5 This is a flowchart of sub-step S23 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application.

[0025] Figure 6This is a flowchart of sub-step S4 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application.

[0026] Figure 7 This is a flowchart of sub-step S5 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application.

[0027] Figure 8 This is a block diagram of an LED display brightness uniformity correction system based on current feedback according to an embodiment of this application. Detailed Implementation

[0028] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0029] To address the problems mentioned above in the background technology, this application proposes a method for correcting the brightness uniformity of LED displays based on current feedback. Figure 1 This is a flowchart of an LED display brightness uniformity correction method based on current feedback according to an embodiment of this application. Figure 2 This is a data flow diagram of an LED display brightness uniformity correction method based on current feedback according to an embodiment of this application. Figure 1 and Figure 2 As shown, the LED display brightness uniformity correction method based on current feedback includes the following steps: S1, in response to the detection of a vertical blanking period synchronization signal, injecting a low-amplitude probe current and a high-amplitude probe current into the target LED pixel to obtain the measured voltage under the low-amplitude probe current and the measured voltage under the high-amplitude probe current; S2, based on the reference IVT characteristic spectrum, performing state decoupling and real-time parameter estimation on the measured voltage under the low-amplitude probe current, the measured voltage under the high-amplitude probe current, the low-amplitude probe current, and the high-amplitude probe current to obtain the estimated real-time junction temperature and the estimated aging factor; S3, extracting the original grayscale value of the target LED pixel from the video data stream; S4, performing dynamic target current calculation on the estimated real-time junction temperature, the estimated aging factor, and the original grayscale value based on the reference LI brightness model to obtain the dynamically adjusted target current; S5, performing closed-loop driving and PWM correction on the original grayscale value and the dynamically adjusted target current to obtain the corrected PWM duty cycle.

[0030] In the aforementioned LED display brightness uniformity correction method based on current feedback, step S1 involves injecting low-amplitude and high-amplitude probe currents into the target LED pixel in response to the detection of a vertical blanking period synchronization signal, thereby obtaining the measured voltage under the low-amplitude probe current and the measured voltage under the high-amplitude probe current. It should be understood that since the pixels are in an emitting state during normal LED display operation, directly injecting probe current would interfere with the image output. Furthermore, the vertical blanking period is a time window between frames where there is no effective display content, and probes performed during this period are not perceptible to the human eye. Therefore, this application further injects high and low amplitude probe currents sequentially into the target LED pixel after detecting the synchronization signal and collects the corresponding voltages to obtain the electrical response data of the LED under different current excitations. This allows for the acquisition of the two-point IV characteristic parameters required for constructing a dynamic resistance model without affecting the normal display effect, providing core data support for the subsequent decoupling of temperature and aging factors, and ensuring the real-time performance and concealment of the correction process.

[0031] In particular, in one specific embodiment, Figure 3 This is a flowchart of sub-step S1 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application. Figure 3 As shown, step S1 includes: S11, generating a probe enable trigger signal in response to detecting a vertical blanking period synchronization signal; S12, injecting a low-amplitude probe current into the target LED pixel in response to the probe enable trigger signal to obtain the measured voltage under the low-amplitude probe current; S13, injecting a high-amplitude probe current into the target LED pixel in response to receiving the measured voltage under the low-amplitude probe current to obtain the measured voltage under the high-amplitude probe current.

[0032] Specifically, in step S11, in response to the detection of the vertical blanking period synchronization signal, a probe enable trigger signal is generated. It should be understood that since the vertical blanking period includes a timing transition segment related to frame switching, directly initiating the probe may lead to data distortion due to circuit instability, and the probe action needs to be precisely synchronized with subsequent current injection, voltage sampling, and other stages. Therefore, this application further generates a standardized probe enable trigger signal after detecting the vertical blanking period synchronization signal to define the effective time window of the probe action and synchronize each execution module. This ensures that the probe process is strictly limited to the stable period of the vertical blanking period, avoiding probe failures caused by timing conflicts, and providing a unified start-up reference for hardware modules such as the current source and ADC, ensuring the coordination and consistency of actions in each stage.

[0033] Specifically, in one possible embodiment, step S11 is implemented as follows: Upon detecting a valid trigger edge of the signal, an internally preset delay counter is immediately started. This counter is used to wait for a stable margin time for the pixel driving circuit to leave the display state of the previous frame. After the delay count ends, a high-level detection enable trigger signal with a fixed pulse width is generated. The pulse width of this signal strictly matches the total duration required for dual-point detection, ensuring that the detection operation is fully executed within the blanking period. Finally, the trigger signal is output to the current generation module and the voltage sampling module.

[0034] Specifically, in step S12, in response to the detection enable trigger signal, a low-amplitude detection current is injected into the target LED pixel to obtain the measured voltage under the low-amplitude detection current. It should be understood that since dual-point detection requires acquiring voltage responses under two different currents, the low-amplitude detection current avoids the instantaneous impact of a large current on the pixel and reflects the electrical characteristics of the pixel under low load conditions, providing basic data for subsequent dynamic resistance calculations. Therefore, this application further prioritizes injecting a low-amplitude detection current and acquiring the corresponding voltage after receiving the detection enable trigger signal, thereby obtaining stable IV data of the pixel under low excitation conditions. In this way, while ensuring the safety of the pixel device, a first set of reliable electrical response data can be obtained, laying the foundation for constructing a dynamic resistance model in conjunction with high-amplitude data, while simultaneously enabling the orderly initiation of the detection process.

[0035] Specifically, in one possible embodiment, step S12 is implemented as follows: First, the precision current source module receives and activates the detection enable trigger signal. Then, it reads the pre-stored low-amplitude detection current configuration word from the internal configuration register and converts it into a corresponding analog current signal via a digital-to-analog converter. Subsequently, the low-amplitude detection current is precisely routed to the driving channel of the target LED pixel via an analog switch matrix. After current injection, a preset current stabilization time is waited to ensure the LED forward voltage reaches a steady state. Then, the high-speed ADC connected to the voltage sampling front end is activated to quantize the voltage across the LED. Finally, the digitized measured voltage under the low-amplitude detection current is stored in a dedicated data register, and a sampling completion signal is sent to the subsequent control module.

[0036] Specifically, in step S13, in response to receiving the measured voltage under a low-amplitude probe current, a high-amplitude probe current is injected into the target LED pixel to obtain the measured voltage under the high-amplitude probe current. It should be understood that since a dynamic resistance reflecting changes in the pixel's electrical characteristics cannot be constructed solely through a low-amplitude probe current, the resistance parameter needs to be extracted through the response difference between high and low amplitude currents. Furthermore, the completion of low-amplitude data sampling indicates that the preceding steps are normal and subsequent detection can be initiated. Therefore, this application further injects a high-amplitude probe current and collects the corresponding voltage immediately after receiving the low-amplitude measured voltage to obtain the IV data of the pixel in a high-excitation state. This allows for rapid acquisition of dual-point detection data, ensuring the temporal correlation of the two sets of data, providing a complete electrical sample for accurate calculation of the dynamic resistance, and ensuring efficient connection of the detection process.

[0037] Specifically, in one possible embodiment, step S13 is implemented as follows: First, a high-amplitude current start command is sent to the precision current source module. Upon receiving the command, the current source module quickly switches its internal configuration, reads the high-amplitude detection current configuration word, and generates a corresponding analog current signal. This high-amplitude current is injected into the target LED pixel through the same driving channel as the low-amplitude detection, ensuring the consistency of the detected object. After the current stabilizes, the high-speed ADC is triggered to perform voltage sampling again. After the ADC completes quantization, the measured voltage corresponding to the high amplitude is stored in another dedicated data register. Finally, the two sets of IV data are packaged and transmitted to the state decoupling and parameter estimation module to complete the entire dual-point detection process.

[0038] In the aforementioned LED display brightness uniformity correction method based on current feedback, step S2, based on the reference IVT characteristic spectrum, performs state decoupling and real-time parameter estimation on the measured voltage under low-amplitude probe current, the measured voltage under high-amplitude probe current, the low-amplitude probe current, and the high-amplitude probe current to obtain the estimated real-time junction temperature and the estimated aging factor. It should be understood that, because the influence of temperature drift and device aging on voltage is coupled in the IV data acquired by dual-point detection, they cannot be directly distinguished. However, the reference IVT characteristic spectrum pre-stores the electrical characteristics of healthy LEDs under different temperatures and currents, which can be used as a decoupling basis. Therefore, this application further relies on this spectrum to process the measured dual sets of IV data to achieve state decoupling and parameter estimation, thereby separating the two key state parameters: real-time junction temperature and aging factor. This overcomes the technical bottleneck of traditional solutions that cannot distinguish the effects of temperature and aging, providing accurate physical state input for subsequent dynamic target current calculations, ensuring that the correction algorithm can specifically compensate for thermal quenching and aging decay, and avoiding brightness deviations caused by miscompensation.

[0039] In particular, in one specific embodiment, Figure 4This is a flowchart of sub-step S2 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application. Figure 4 As shown, step S2 includes: S21, determining the dynamic resistance value based on the measured voltage under low-amplitude probe current, the measured voltage under high-amplitude probe current, the low-amplitude probe current, and the high-amplitude probe current; S22, estimating the junction temperature of the dynamic resistance value based on the dynamic resistance-junction temperature relationship model to obtain the estimated real-time junction temperature; S23, estimating the aging factor based on the reference IVT characteristic spectrum, the measured voltage under low-amplitude probe current, the low-amplitude probe current, and the estimated real-time junction temperature to obtain the estimated aging factor.

[0040] Specifically, step S21 determines the dynamic resistance value based on the measured voltage under low-amplitude probe current, the measured voltage under high-amplitude probe current, the low-amplitude probe current, and the high-amplitude probe current. It should be understood that since the dynamic resistance of an LED directly reflects its junction temperature change, and this parameter needs to be extracted through the voltage response differences under different currents, a single set of IV data cannot construct a dynamic resistance model and cannot accurately characterize the changes in the LED's electrical state. Therefore, this application further utilizes two sets of IV data obtained through dual-point detection, and determines the dynamic resistance value by calculating the ratio of the voltage difference to the current difference, thereby obtaining the core electrical parameter that can correlate with the LED's temperature state. In a specific example of this application, step S21 includes: determining the dynamic resistance value using the following formula:

[0041]

[0042] in, The measured voltage under high amplitude probe current, For the measured voltage under low amplitude probe current, For high-amplitude detection current, For low-amplitude detection current, This is a dynamic resistance value. This allows for accurate capture of the dynamic electrical characteristics of the LED under its current operating state, providing a reliable data source for subsequent resistance-based junction temperature estimation. It avoids resistance calculation errors caused by single sets of data, improving the accuracy of junction temperature estimation from the source.

[0043] Specifically, in step S22, based on the dynamic resistance-junction temperature relationship model, the junction temperature is estimated from the dynamic resistance value to obtain the estimated real-time junction temperature. It should be understood that since there is a clear and stable physical correlation between the dynamic resistance and junction temperature of an LED, an increase in junction temperature will cause the dynamic resistance to change systematically. This correlation has been stored in the dynamic resistance-junction temperature relationship model through offline calibration, and the model can be used to reversely deduce from resistance to junction temperature. Therefore, this application further inputs the calculated dynamic resistance value into the model, and estimates the junction temperature of the dynamic resistance through model calculations to obtain the current real-time junction temperature of the LED pixel. In this way, junction temperature detection can be achieved directly using the LED's own electrical characteristics without the need for an additional integrated temperature sensor. This reduces hardware costs and layout complexity, avoids the coarse granularity problem of module-level temperature measurement, and ensures the accuracy and real-time performance of the junction temperature estimation for each pixel.

[0044] Specifically, in one possible embodiment, step S22 is implemented as follows: First, the dynamic resistance value is read from the intermediate register, and the dynamic resistance-junction temperature relationship model stored in the non-volatile memory is invoked. If the model is in the form of a lookup table, the dynamic resistance value is used as an index to locate two adjacent resistance-junction temperature data pairs in the table, and then a precise real-time junction temperature is obtained through linear interpolation. If the model is in the form of a polynomial function, the dynamic resistance value is substituted into a pre-stored quadratic polynomial formula, and multiplication and addition operations are performed to obtain the real-time junction temperature. After the calculation is completed, the result is checked for range to ensure that it is within the normal operating temperature range of the LED. After the check passes, the real-time junction temperature is stored in the result register.

[0045] Specifically, in step S23, based on the reference IVT characteristic spectrum, an aging factor is estimated by analyzing the measured voltage under low-amplitude probe current, the low-amplitude probe current, and the estimated real-time junction temperature to obtain the estimated aging factor. It should be understood that, after excluding the influence of junction temperature, the deviation between the measured LED voltage and the theoretical voltage under healthy conditions is mainly caused by irreversible aging. Aging increases the LED series resistance and intensifies non-radiative recombination, leading to a forward voltage shift. The reference IVT characteristic spectrum provides a healthy voltage benchmark under different junction temperatures and currents. Therefore, this application further uses real-time junction temperature and low-amplitude probe current as query conditions to obtain the theoretical voltage from the reference spectrum, and then calculates the aging factor by the deviation between the measured voltage and the theoretical voltage, thereby achieving a quantitative estimate of the aging degree. This completely eliminates the interference of temperature on voltage, accurately locates the device characteristic drift caused by aging, provides accurate parameters for subsequent current compensation for aging, and avoids overcompensation or undercompensation caused by misjudging temperature effects as aging.

[0046] In particular, in one specific embodiment, Figure 5This is a flowchart of sub-step S23 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application. Figure 5 As shown, step S23 includes: S231, querying the reference IVT characteristic spectrum by inputting the low-amplitude probe current and the estimated real-time junction temperature to obtain the theoretical forward voltage; S232, calculating the voltage difference between the theoretical forward voltage and the measured voltage under the low-amplitude probe current to obtain the voltage bias; S233, inputting the voltage bias into a preset conversion function to obtain the estimated aging factor.

[0047] More specifically, in step S231, the low-amplitude probe current and the estimated real-time junction temperature are input into a reference IVT characteristic spectrum for querying to obtain the theoretical forward voltage. It should be understood that since the measured voltage alone cannot isolate the effects of aging, it is necessary to obtain the electrical characteristics of the LED under the current junction temperature and current in a non-aging state as a reference. The reference IVT characteristic spectrum fully records the current-voltage-temperature correlation of a healthy LED and can provide this reference data. Therefore, this application further uses the low-amplitude probe current and the estimated junction temperature as a dual-dimensional index, inputting them into the reference IVT characteristic spectrum for querying, thereby obtaining the theoretical forward voltage excluding aging interference. This establishes an accurate benchmark for subsequent quantification of voltage deviation caused by aging, avoiding the inability to distinguish the source of voltage change due to a lack of healthy state reference, ensuring the accuracy of aging factor calculation from the data source, and providing key support for decoupling the effects of temperature and aging.

[0048] Specifically, in one possible embodiment, step S231 is implemented as follows: First, the real-time junction temperature and low-amplitude probe current are received. Then, a reference IVT characteristic spectrum stored in a non-volatile memory is invoked. This spectrum stores discrete theoretical voltage values ​​with junction temperature and current as two-dimensional axes. Next, a two-dimensional indexing is performed with the real-time junction temperature as the Y-axis and the low-amplitude probe current as the X-axis. If the index value falls on a grid point, the voltage is read directly; if it falls on a gap, a linear interpolation unit is activated to calculate the accurate theoretical voltage based on four adjacent data points. After calculation, the voltage verification module confirms that the result is within the normal operating range of the LED. After successful verification, the theoretical voltage is stored in an intermediate register, and the deviation calculation module is triggered.

[0049] More specifically, step S232 calculates the voltage difference between the theoretical forward voltage and the measured voltage under low-amplitude probe current to obtain the voltage bias. It should be understood that since the theoretical forward voltage represents the electrical state of a healthy LED, while the measured voltage includes characteristic drift caused by aging, the impact of aging cannot be quantified by a single voltage value; the difference between the healthy and aging states needs to be separated. Therefore, this application further performs a difference calculation on the theoretical forward voltage and the measured voltage under low-amplitude probe current to obtain the voltage bias caused solely by irreversible aging. This transforms the impact of aging on the LED's electrical characteristics into a quantifiable parameter, eliminates interference from the health benchmark, provides direct raw data for subsequently converting the voltage offset into an aging factor, avoids erroneous adjustments in subsequent compensation algorithms due to input errors, and ensures the accuracy of aging assessment.

[0050] Specifically, in one possible embodiment, step S232 is implemented as follows: First, the theoretical forward voltage is read from the intermediate register, and the low-amplitude measured voltage is read from the probe data register. Next, the internal difference calculator performs the operation of subtracting the theoretical voltage from the measured voltage to obtain the initial deviation. Then, the initial deviation is input to the noise suppression module, and the ADC sampling noise and current fluctuation error are eliminated through moving average filtering. After filtering, the deviation range judgment module is activated. If the deviation exceeds the reasonable range of LED aging voltage offset, the process of re-querying the theoretical voltage and the sampled measured voltage is triggered; if it is within the reasonable range, the final voltage bias is stored in the deviation register, and a deviation ready signal is sent to the aging factor conversion module.

[0051] More specifically, in step S233, the voltage bias is input into a preset conversion function to obtain an estimated aging factor. It should be understood that since the voltage bias is an electrical offset, it cannot directly adapt to the parameter requirements of the subsequent dynamic current compensation algorithm. It needs to be mapped to a normalized index characterizing the degree of aging through a preset conversion relationship, and this conversion relationship, after offline calibration, can match the aging characteristics of LEDs. Therefore, this application further inputs the voltage bias into a preset conversion function to perform calculations, thereby obtaining an estimated aging factor that can be directly used for compensation calculations. In this way, the abstract voltage offset can be transformed into a concrete aging degree parameter, enabling the subsequent dynamic target current calculation to accurately match the luminous efficacy decay caused by aging, avoiding compensation deviations caused by incompatible parameter forms, and ensuring the consistency and accuracy of brightness correction.

[0052] Specifically, in one possible embodiment, step S233 is implemented as follows: First, the voltage bias is read from the bias register, and simultaneously, a pre-stored conversion function is retrieved. This function is a linear formula or lookup table calibrated through an offline accelerated aging experiment. The calibration process includes: placing the sample LED in a high-temperature (e.g., 85°C) and overdriven (e.g., 1.2 times the rated current) environment for continuous operation for 1000 to 3000 hours, and periodically collecting the correspondence between voltage drift and light decay data for fitting. If it is a linear function, the voltage bias is substituted into "aging factor = voltage bias × proportional coefficient + intercept" for calculation, where the proportional coefficient is usually 0.02 to 0.05 (unit: 1 / mV), and the intercept is usually 0; if it is a lookup table, the voltage bias is used as the index for location, and if it is not at a discrete point, nearest neighbor interpolation is performed. After obtaining the initial aging factor, it is smoothed by a first-order low-pass filter to smooth instantaneous fluctuations. Finally, the filtered aging factor is stored in the aging parameter register, a ready signal is sent to the dynamic target current calculation module, and the long-term aging data log of the pixel is updated.

[0053] In the aforementioned LED display brightness uniformity correction method based on current feedback, step S3 involves extracting the original grayscale value of the target LED pixel from the video data stream. It should be understood that dynamic target current calculation requires specifying the brightness command that the pixel should display in the current frame. The video data stream carries the pixel brightness information of each frame, and the original grayscale value is the digital representation of this information, serving as the source of the subsequent brightness correction command. Therefore, this application further extracts the original grayscale value corresponding to the target LED pixel from the input video data stream to obtain the basic brightness display command for that pixel. This provides a core command input for subsequent calculation of the dynamic target current based on junction temperature and aging factor, ensuring that the current adjustment accurately matches the pixel's display requirements. This avoids deviation of the correction direction from the actual display requirements due to a lack of brightness commands, ensuring consistency between the corrected brightness and the image content.

[0054] Specifically, in one possible embodiment, step S3 is implemented as follows: First, a video data stream from the display screen's main controller is received. This data stream follows a preset video format, such as HDMI or DP protocols. The protocol parsing unit inside the module decodes the data stream, separating the frame synchronization signal, the line synchronization signal, and the pixel data. Then, based on the row and column coordinates of the target LED pixel on the display screen, the digital grayscale value of the pixel is captured within the corresponding row and column timing window during each frame of data transmission. The captured grayscale value is converted into an internally unified 10-bit or 12-bit digital format to avoid data loss due to format incompatibility. Finally, the converted original grayscale value is stored in a dedicated pixel data register, and a grayscale value ready signal is sent to the dynamic target current calculation module.

[0055] In the aforementioned LED display brightness uniformity correction method based on current feedback, step S4 involves dynamically calculating the target current based on the estimated real-time junction temperature, estimated aging factor, and original grayscale value using a reference LI brightness model to obtain a dynamically adjusted target current. It should be understood that since LED brightness output is affected by both junction temperature rise and aging degradation, a fixed drive current will cause the actual brightness to deviate from the target value. The reference LI brightness model records the brightness-current relationship of a healthy LED, which can be corrected by combining junction temperature and aging factor. Therefore, this application further relies on the reference LI brightness model, integrating the estimated real-time junction temperature, aging factor, and original grayscale value to perform dynamic target current calculation, thereby obtaining a drive current that can offset thermal quenching and aging degradation. This ensures that under different temperatures and aging conditions, the actual brightness of the LED accurately matches the target brightness corresponding to the original grayscale value through dynamic current adjustment, avoiding the degradation of full-screen uniformity caused by accumulated brightness deviations, and significantly improving the dynamic adaptability of the correction.

[0056] In particular, in one specific embodiment, Figure 6 This is a flowchart of sub-step S4 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application. Figure 6 As shown, step S4 includes: S41, processing the original grayscale value based on the gamma correction lookup table to obtain the target brightness; S42, determining the comprehensive brightness efficiency factor based on the estimated real-time junction temperature and the estimated aging factor; S43, dividing the target brightness by the comprehensive brightness efficiency factor to obtain the compensated brightness target; S44, inputting the compensated brightness target into the reference LI brightness model to obtain the dynamically adjusted target current.

[0057] Specifically, step S41 involves processing the original grayscale values ​​using a gamma correction lookup table to obtain the target brightness. It should be understood that because human vision perceives brightness non-linearly, linearly mapping the original grayscale values ​​to brightness would result in loss of detail in low grayscale ranges and saturation in high grayscale ranges. The gamma correction lookup table pre-stores non-linear mapping relationships that conform to human visual perception. Therefore, this application further utilizes the gamma correction lookup table to perform non-linear transformation processing on the original grayscale values ​​to obtain a target brightness that matches the human visual system. This corrects the non-linear distortion of brightness perception, ensuring clear details in low grayscale ranges and natural brightness in high grayscale ranges, avoiding the loss of image depth caused by linear mapping, and significantly improving the visual realism and comfort of the displayed image.

[0058] Specifically, in one possible embodiment, step S41 is implemented as follows: First, the original grayscale value is read from the pixel data register, and a pre-calibrated gamma correction lookup table is retrieved from non-volatile memory. This lookup table uses the original grayscale value as an index and stores corresponding linear brightness weight values, which are optimized offline to match human visual characteristics. If the read original grayscale value falls exactly on a discrete index point of the lookup table, the corresponding linear brightness weight is directly read; if it falls within an index gap, the precise weight is calculated based on the weight values ​​of two adjacent index points. Then, the linear brightness weight is multiplied by the preset peak brightness specification of the display screen to obtain the absolute physical target brightness corresponding to the current grayscale value.

[0059] Specifically, in step S42, a comprehensive luminance efficiency factor is determined based on the estimated real-time junction temperature and the estimated aging factor. It should be understood that since the luminous efficiency of an LED decreases with increasing junction temperature and also decays with increasing aging, and these two factors have a coupled effect on efficiency (aged LEDs are more sensitive to temperature), considering either factor alone can lead to errors in efficiency assessment. Therefore, this application further combines the estimated real-time junction temperature and the aging factor to determine the comprehensive luminance efficiency factor through a preset joint efficiency model, thereby quantifying the actual luminous efficiency of the LED under the current state. This accurately characterizes the degree of efficiency decay under the coupled effect of temperature and aging, avoiding inaccurate efficiency parameters caused by separate assessments, providing a reliable basis for subsequent calculations of the compensation luminance target, and ensuring that the dynamic target current can accurately offset the efficiency decay.

[0060] Specifically, in one possible embodiment, step S42 is implemented as follows: First, the estimated real-time junction temperature and aging factor are read from the result register, and a pre-stored two-dimensional joint efficiency model is retrieved. This model is calibrated through offline accelerated aging experiments and stores efficiency values ​​with junction temperature and aging factor as axes. The model is positioned in the discrete grid of the joint efficiency model with real-time junction temperature as the vertical coordinate and aging factor as the horizontal coordinate. If the coordinate falls on a grid point, the corresponding efficiency factor is directly read; if it falls in a grid gap, a precise efficiency value is obtained through bilinear interpolation. The calculated efficiency factor is then input into a range verification module to confirm that it is within a reasonable range of 0.1-1.0, thus eliminating abnormal data.

[0061] In particular, in another preferred embodiment, when determining the overall luminance efficiency factor based on the estimated real-time junction temperature and the estimated aging factor, since the temperature efficiency term and the aging efficiency term are not completely independent, that is, the effect of temperature on LED luminance efficiency is not completely independent of the aging state of the LED, it is necessary to optimize the way they affect the overall efficiency.

[0062] Specifically, the aging process of LEDs (e.g., the increase of internal defects) may alter their sensitivity to temperature changes. The efficiency decline of a healthy new LED at rising temperatures may differ from that of an LED that has already undergone some degree of aging at the same temperature increase. In other words, aging may cause the brightness decay of LEDs to be more severe at high temperatures, or vice versa. Furthermore, the photoelectric conversion efficiency of brightness is a complex physical process, influenced by the nonlinear superposition of various factors such as semiconductor material characteristics, junction temperature, injection current density, and crystal defects. Moreover, the effects of temperature and aging on these microscopic mechanisms also exhibit a complex nonlinear interaction.

[0063] Therefore, in order to capture this complex physical phenomenon more accurately, the estimated real-time junction temperature and the estimated aging factor should be used as joint inputs to query a preset two-dimensional joint feature model to obtain the comprehensive brightness efficiency factor. For example, a two-dimensional or multi-dimensional joint feature model can be established to achieve this, which can inherently capture any correlation or nonlinear interaction between temperature and aging.

[0064] First, in the offline calibration phase, a benchmark model is established. Specifically, a two-dimensional polynomial surface model can be constructed to perform a complete LI characteristic scan on LED samples with different aging levels (obtained through prior accelerated aging experiments) at multiple different ambient temperatures, obtaining a series of data points. Then, these data points are used to fit a two-dimensional model. For example, the fitted two-dimensional polynomial surface model is represented as:

[0065]

[0066] in, It is an estimated real-time junction temperature. These are estimated aging factors. These are the coefficients of the model, solved through offline fitting, and It is the highest order of the polynomial, which determines the complexity and fitting ability of the model. It is the comprehensive brightness efficiency factor. Based on a fitted two-dimensional polynomial surface model, it can be customized for typical data and a 2D-LUT lookup table can be used.

[0067] Then, for the estimated real-time junction temperature Estimated aging factors These two values ​​are used as two-dimensional coordinates to perform a query within a pre-stored two-dimensional model. Here, because... and These are continuous, real-time calculated values, which may not fall precisely on the 2D-LUT mesh points of the 2D model. Therefore, bilinear interpolation is required to calculate the accurate overall brightness efficiency factor. .

[0068] Here, when the estimated real-time junction temperature and the estimated aging factor fall within the region defined by the four grid points (T1, A1), (T2, A1), (T1, A2), and (T2, A2) in the two-dimensional joint feature model: the lower left corner ( , The efficiency value is bottom right corner ( , The efficiency value is Top left corner ( , The efficiency value is Top right corner ( , The efficiency value is Then, bilinear interpolation is performed to calculate the overall luminance efficiency factor. Bilinear interpolation includes: first, performing linear interpolation along the temperature axis to calculate the first temporary efficiency value and the second temporary efficiency value corresponding to the estimated real-time junction temperature at the first aging level A1 and the second aging level A2, respectively. That is, calculating the efficiency at the aging level A1... At that time, the temperature was Temporary efficiency value :

[0069]

[0070] And calculate at the aging level of At that time, the temperature was Temporary efficiency value :

[0071]

[0072] Then, based on the first and second temporary efficiency values, linear interpolation is performed along the aging axis to calculate the overall luminance efficiency factor. :

[0073]

[0074] Thus, because the two-dimensional model can accurately characterize the coupling effect of temperature and aging, the calculated... By more closely approximating the true physical efficiency of LEDs, the calculated compensated brightness target becomes more accurate, resulting in more precise brightness control at the hardware level. Consequently, throughout the entire lifespan of the display, especially in the later stages of aging, there is no accumulation or amplification of errors. It can continuously and accurately compensate pixels under different aging conditions, significantly improving the uniformity of brightness and color throughout the entire lifecycle of the display after prolonged use.

[0075] Furthermore, for applications with large operating temperature differences, such as outdoor displays, it can provide appropriate temperature compensation for aging pixels, whether in the cold winter morning or the hot summer afternoon, avoiding the phenomenon of being too bright when cold and too dark when hot, or vice versa, thus having stronger wide temperature range adaptability.

[0076] Specifically, in step S43, the target brightness is divided by the comprehensive brightness efficiency factor to obtain the compensated brightness target. It should be understood that since the comprehensive brightness efficiency factor reflects the degree to which the actual luminous efficiency of the LED is below its healthy state, if the target brightness is used directly for driving, the actual output brightness will be lower than expected due to efficiency degradation. Therefore, the required "virtual" brightness target needs to be calculated by working backwards from the efficiency degradation ratio. Thus, this application further performs a division operation between the target brightness and the comprehensive brightness efficiency factor to obtain a compensated brightness target that can offset the efficiency degradation. This ensures that the driving current calculated based on this compensation target can still make the LED output meet the expected actual brightness under the current low-efficiency state, avoiding insufficient brightness due to efficiency degradation and providing an accurate brightness compensation benchmark for dynamic target current calculation.

[0077] Specifically, in one possible embodiment, step S43 is implemented as follows: First, the target brightness is read from the brightness register, and the comprehensive brightness efficiency factor is read from the efficiency register. The division unit inside the unit performs the operation of dividing the target brightness by the comprehensive brightness efficiency factor to obtain the initial compensated brightness target. Considering the numerical error that may be introduced by the division operation, the error correction module is activated to calibrate the initial result using a pre-stored correction coefficient. Subsequently, the calibrated compensated brightness target is input into the brightness range limiting module to ensure that it does not exceed the maximum physical luminous brightness of the LED and avoid over-driving. If it exceeds the range, it is clamped to the maximum brightness value. Finally, the compliant compensated brightness target is stored in the compensated brightness register.

[0078] Specifically, in step S44, the compensated brightness target is input into the reference LI brightness model to obtain the dynamically adjusted target current. It should be understood that since the reference LI brightness model records the correspondence between the brightness and driving current of a healthy LED under ideal conditions, the compensated brightness target is the "virtual" brightness required to offset efficiency degradation, and the corresponding driving current needs to be deduced from this model. Therefore, this application further inputs the compensated brightness target into the inverse model of the reference LI brightness model to obtain the dynamically adjusted target current that enables the LED to output the expected brightness. This allows for precise matching of the degree of efficiency degradation, ensuring that the driving current is completely adapted to the current physical state of the LED, avoiding brightness deviations caused by fixed currents, and ultimately achieving precise brightness control under different temperatures and aging conditions.

[0079] Specifically, in one possible embodiment, step S44 is implemented as follows: First, the compensated brightness target is read from the compensated brightness register, and the pre-stored reference LI brightness model is retrieved in the form of a reverse lookup table with brightness as the index and current as the output. The compensated brightness target is used as the index to locate the value in the reverse lookup table. If the index value falls on a discrete data point, the corresponding drive current value is directly read. If it falls within a data point gap, the linear interpolation unit is activated to calculate the precise current value based on two adjacent brightness-current data pairs. After calculating the initial target current, it is input to the current calibration module, which fine-tunes the current value based on the forward voltage characteristics of the LED to eliminate the influence of voltage fluctuations. Finally, the calibrated dynamically adjusted target current is stored in the drive current register, and a current setting ready signal is sent to the closed-loop drive and PWM correction module.

[0080] In the aforementioned LED display brightness uniformity correction method based on current feedback, step S5 involves closed-loop driving and PWM correction of the original grayscale value and the dynamically adjusted target current to obtain the corrected PWM duty cycle. It should be understood that since the dynamically adjusted target current is a theoretically calculated value, in actual driving, it is easily affected by power supply voltage fluctuations, driver transistor parameter drift, and other interferences, causing the actual current to deviate from the target value. Furthermore, the original grayscale value needs to be converted into a PWM signal to drive the LED, requiring dynamic PWM correction based on errors. Therefore, this application further integrates the original grayscale value and the dynamic target current, generating the final duty cycle through a closed-loop driving and PWM correction process to ensure that the actual driving current accurately tracks the target value. This allows for real-time cancellation of external interference and the effects of non-ideal circuit characteristics, avoiding brightness fluctuations caused by deviations between the theoretical current and the actual driving current, ensuring that the LED output brightness always matches the target brightness, and significantly improving the robustness of the correction.

[0081] In particular, in one specific embodiment, Figure 7 This is a flowchart of sub-step S5 of the LED display brightness uniformity correction method based on current feedback according to an embodiment of this application. Figure 7 As shown, step S5 includes: S51, performing real-time current sampling and error calculation on the dynamically adjusted target current to obtain a current error signal; S52, generating a reference PWM duty cycle based on the original grayscale value; S53, inputting the current error signal into the PI controller to obtain the duty cycle correction amount; S54, synthesizing and saturating the reference PWM duty cycle and the duty cycle correction amount to obtain the corrected PWM duty cycle.

[0082] Specifically, in step S51, the dynamically adjusted target current is sampled and its error is calculated in real time to obtain a current error signal. It should be understood that since the dynamic target current is a theoretically set value, during actual driving, factors such as power supply voltage fluctuations and driver chip temperature drift can cause the actual current to deviate from the target value. If the deviation is not monitored in real time, closed-loop correction will lose its basis. Therefore, this application further samples the actual driving current corresponding to the dynamic target current in real time and calculates the difference between the two to obtain an error signal reflecting the current deviation. This provides a precise deviation input for the subsequent PI controller, ensuring that the correction amount perfectly matches the current deviation, offsetting the effects of external interference and circuit drift in real time, avoiding brightness deviation caused by the actual current deviating from the target, and providing core feedback for closed-loop correction.

[0083] Specifically, in one possible embodiment, step S51 is implemented as follows: First, the dynamically adjusted target current is received, and the PWM turn-on synchronization signal is acquired. During the conduction period when the PWM signal is at a high level, the unit controls the current sensing amplifier to amplify the voltage signal across the series resistor to eliminate noise interference. Subsequently, the high-speed ADC quantizes the amplified voltage signal to obtain the digital value of the actual current. Then, the dynamic target current is subtracted from the actual current to obtain a signed current error signal. To avoid the influence of instantaneous noise, the error signal is input to a first-order low-pass filter module to smooth fluctuations. Finally, the filtered current error signal is stored in the error register, and an error ready signal is sent to the PI controller.

[0084] Specifically, in step S52, a reference PWM duty cycle is generated based on the original grayscale value. It should be understood that since the LED needs to adjust its brightness by controlling the conduction time through a PWM signal, the original grayscale value, as a brightness command, must first be converted into a basic PWM duty cycle. This duty cycle serves as the reference for subsequent error correction; without a reference value, the correction amount will have no reference object. Therefore, this application further generates a reference PWM duty cycle based on the original grayscale value through a preset mapping relationship, thereby establishing a basic drive signal that matches the original brightness command. This ensures that the basic duty cycle of the PWM signal is consistent with the brightness requirement corresponding to the original grayscale value. Subsequent correction only requires fine-tuning the duty cycle to offset deviations, avoiding deviations in the correction direction due to the lack of a reference, and ensuring that brightness adjustment always revolves around the target value.

[0085] Specifically, in one possible embodiment, step S52 is implemented as follows: First, the original grayscale value is read from the pixel data register. The unit internally retrieves a pre-stored grayscale-PWM mapping table, which is established through offline calibration, mapping different grayscale values ​​to PWM duty cycles of 0%-100%, and the mapping relationship conforms to the linear adjustment requirements of LED brightness. If the original grayscale value falls exactly on a discrete index point in the mapping table, the corresponding reference duty cycle is directly read; if it falls within an index gap, the linear interpolation unit is activated to calculate a precise reference value based on the duty cycles corresponding to two adjacent grayscale values. Subsequently, the reference duty cycle is converted into a digital format supported by the driver chip, such as a 12-bit count value, stored in the reference PWM register, and a reference ready signal is sent to the synthesis unit.

[0086] Specifically, in step S53, the current error signal is input to the PI controller to obtain the duty cycle correction amount. It should be understood that since static errors cannot be eliminated by proportional control (P control) alone, and the current error needs to be converted into an adjustment amount of the PWM duty cycle to achieve correction, the PI controller combines the characteristics of fast proportional response and integral elimination of static error, enabling precise conversion of deviations. Therefore, this application further inputs the current error signal to a preset PI controller, generating a duty cycle correction amount through calculation, thereby obtaining a PWM adjustment value that matches the current deviation. This allows for rapid response to instantaneous current deviations while completely eliminating long-term static errors, avoiding continuous current deviations caused by control algorithm defects, ensuring stable tracking of the dynamic target current by the actual current, and providing algorithmic support for brightness stability.

[0087] Specifically, in one possible embodiment, step S53 is implemented as follows: First, the current error signal is read from the error register, and the pre-stored proportional coefficient (Kp) and integral coefficient (Ki) are retrieved. The value range of Kp is set to 0.3 to 0.8 to balance the current response speed and overshoot; the value range of Ki is set to 0.01 to 0.05 to avoid system oscillation while eliminating steady-state error. Next, the current error is multiplied by Kp to obtain the proportional term correction. Then, the current error is accumulated and integrated, and multiplied by Ki to obtain the integral term correction. Subsequently, the adder unit superimposes the proportional term and the integral term to obtain the initial duty cycle correction. To prevent overshoot caused by integral saturation, the integral limiting unit clamps the integral term within a preset range. Finally, the processed duty cycle correction is stored in the correction amount register, and a correction amount ready signal is sent to the synthesis unit, while resetting the next round of integral accumulator.

[0088] Specifically, in step S54, the reference PWM duty cycle and the duty cycle correction amount are synthesized and saturated to obtain the corrected PWM duty cycle. It should be understood that since the reference PWM duty cycle corresponds to the basic brightness requirement of the original grayscale value, and the duty cycle correction amount is used to offset current deviations, both need to be superimposed to obtain an accurate drive signal. Furthermore, the superimposed value may exceed the effective range of the PWM, requiring limitation to protect the LED. Therefore, this application further synthesizes the reference duty cycle and the correction amount, and limits the effective range of the result to obtain a compliant final PWM duty cycle. This allows for real-time correction of current deviations while meeting basic brightness requirements, preventing the synthesized value from exceeding the range and causing LED overcurrent damage or insufficient brightness, ensuring both accurate and safe drive signals, and guaranteeing the correction effect and device reliability.

[0089] Specifically, in one possible embodiment, step S54 is implemented as follows: First, the reference duty cycle is read from the reference PWM register, and the duty cycle correction value is read from the correction value register. The adder adds the two values ​​to obtain the initial corrected PWM duty cycle. Then, the range judgment unit detects whether the initial value is within the valid digital count value range (i.e., 0 to the maximum count value supported by the driver chip). If the initial value is greater than the maximum count value (corresponding to 100% duty cycle), the saturation unit clamps it to the maximum count value to prevent the LED from being overdriven and burned out; if it is less than 0, it is clamped to 0 to avoid abnormally low brightness. The duty cycle after saturation processing is used as the final digital signal output. Finally, the corrected PWM duty cycle is output to the gate driver of the LED driver module, and a drive ready signal is sent to the closed-loop control unit to trigger the next round of sampling correction.

[0090] In summary, the current feedback-based LED display brightness uniformity correction method according to the embodiments of this application is explained. First, it utilizes a vertical blanking period timing window to inject high and low amplitude probe currents into the LED pixels, acquiring the voltage response under different excitation states without interfering with normal display. Then, a state decoupling model based on a reference IVT characteristic spectrum is constructed. Utilizing the differential sensitivity of voltage to temperature and aging, the coupled electrical signals are deeply analyzed into independent real-time junction temperature and aging factors. Furthermore, based on these physical state parameters, inverse compensation calculations are performed using the original grayscale command and the reference LI brightness model to dynamically derive the target driving current capable of offsetting thermal quenching and aging degradation. This effectively solves the correction problem caused by the deep coupling of temperature drift and device aging, thereby achieving adaptive high-precision brightness uniformity correction throughout the entire lifecycle of the display.

[0091] Figure 8 This is a block diagram of an LED display brightness uniformity correction system based on current feedback according to an embodiment of this application. Figure 8As shown, the LED display brightness uniformity correction system 100 based on current feedback according to an embodiment of this application includes: a measured voltage acquisition module 110, used to inject a low-amplitude probe current and a high-amplitude probe current into the target LED pixel in response to the detection of a vertical blanking period synchronization signal to obtain the measured voltage under the low-amplitude probe current and the measured voltage under the high-amplitude probe current; and a state decoupling and parameter estimation module 120, used to perform state decoupling and parameter estimation on the measured voltage under the low-amplitude probe current, the measured voltage under the high-amplitude probe current, and the low-amplitude probe current and the high-amplitude probe current based on a reference IVT characteristic spectrum. The system includes a line state decoupling and real-time parameter estimation module to obtain the estimated real-time junction temperature and the estimated aging factor; a raw gray value extraction module 130 for extracting the raw gray value of the target LED pixel from the video data stream; a dynamic target current calculation module 140 for calculating the dynamic target current based on the estimated real-time junction temperature, the estimated aging factor, and the raw gray value using a reference LI brightness model to obtain the dynamically adjusted target current; and a closed-loop drive and PWM correction module 150 for performing closed-loop drive and PWM correction on the raw gray value and the dynamically adjusted target current to obtain the corrected PWM duty cycle.

[0092] As described above, the LED display brightness uniformity correction system 100 based on current feedback according to the embodiments of this application can be implemented in various wireless terminals, such as servers with an LED display brightness uniformity correction algorithm based on current feedback. In one possible implementation, the LED display brightness uniformity correction system 100 based on current feedback according to the embodiments of this application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the LED display brightness uniformity correction system 100 based on current feedback can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the LED display brightness uniformity correction system 100 based on current feedback can also be one of many hardware modules of the wireless terminal.

[0093] Alternatively, in another example, the LED display brightness uniformity correction system 100 based on current feedback and the wireless terminal can also be separate devices, and the LED display brightness uniformity correction system 100 based on current feedback can be connected to the wireless terminal via wired and / or wireless networks, and transmit interactive information in accordance with an agreed data format.

[0094] Here, those skilled in the art will understand that the specific operations of each step in the above-described LED display brightness uniformity correction system based on current feedback have been referenced above. Figures 1 to 7 The method for correcting the brightness uniformity of LED displays based on current feedback has been described in detail, and therefore, its repeated description will be omitted.

Claims

1. A method for correcting the brightness uniformity of an LED display screen based on current feedback, characterized in that, include: In response to the detection of the vertical blanking period synchronization signal, a low-amplitude probe current and a high-amplitude probe current are injected into the target LED pixel to obtain the measured voltage under the low-amplitude probe current and the measured voltage under the high-amplitude probe current. Based on the baseline IVT characteristic spectrum, state decoupling and real-time parameter estimation are performed on the measured voltage under low amplitude probe current, measured voltage under high amplitude probe current, low amplitude probe current, and high amplitude probe current to obtain the estimated real-time junction temperature and estimated aging factor. Extract the raw grayscale values ​​of the target LED pixels from the video data stream; Based on the baseline LI brightness model, the estimated real-time junction temperature, the estimated aging factor, and the original gray value are used to dynamically calculate the target current to obtain the dynamically adjusted target current. Closed-loop drive and PWM correction are performed on the original grayscale value and the dynamically adjusted target current to obtain the corrected PWM duty cycle.

2. The LED display brightness uniformity correction method based on current feedback according to claim 1, characterized in that, In response to the detection of a vertical blanking synchronization signal, a low-amplitude probe current and a high-amplitude probe current are injected into the target LED pixel to obtain the measured voltage under the low-amplitude probe current and the measured voltage under the high-amplitude probe current, including: In response to the detection of a vertical blanking period synchronization signal, a probe enable trigger signal is generated; In response to the detection enable trigger signal, a low-amplitude detection current is injected into the target LED pixel to obtain the measured voltage under the low-amplitude detection current; In response to receiving the measured voltage under a low-amplitude probe current, a high-amplitude probe current is injected into the target LED pixel to obtain the measured voltage under a high-amplitude probe current.

3. The method for correcting the brightness uniformity of an LED display screen based on current feedback according to claim 1, characterized in that, Based on the baseline IVT characteristic spectrum, state decoupling and real-time parameter estimation are performed on the measured voltage under low-amplitude probe current, measured voltage under high-amplitude probe current, and low-amplitude and high-amplitude probe currents to obtain the estimated real-time junction temperature and estimated aging factor, including: The dynamic resistance value is determined based on the measured voltage under low amplitude probe current, the measured voltage under high amplitude probe current, the low amplitude probe current, and the high amplitude probe current. Based on the dynamic resistance-junction temperature relationship model, the junction temperature is estimated by the dynamic resistance value to obtain the estimated real-time junction temperature; Based on the baseline IVT characteristic spectrum, the aging factor is estimated by analyzing the measured voltage, low-amplitude probe current, and estimated real-time junction temperature under low-amplitude probe current.

4. The LED display brightness uniformity correction method based on current feedback according to claim 3, characterized in that, The dynamic resistance value is determined based on the measured voltage under low-amplitude probe current, the measured voltage under high-amplitude probe current, the low-amplitude probe current, and the high-amplitude probe current. This includes determining the dynamic resistance value using the following formula: in, The measured voltage under high amplitude probe current, For the measured voltage under low amplitude probe current, For high-amplitude detection current, For low-amplitude detection current, This is the dynamic resistance value.

5. The LED display brightness uniformity correction method based on current feedback according to claim 3, characterized in that, Based on the baseline IVT characteristic spectrum, the aging factor is estimated by analyzing the measured voltage, low-amplitude probe current, and estimated real-time junction temperature under low-amplitude probe current, including: The low-amplitude probe current and estimated real-time junction temperature are used to query the reference IVT characteristic spectrum to obtain the theoretical forward voltage; The voltage bias is obtained by calculating the voltage difference between the theoretical forward voltage and the measured voltage under low-amplitude probe current. Input the voltage bias into a preset conversion function to obtain the estimated aging factor.

6. The method for correcting the brightness uniformity of an LED display screen based on current feedback according to claim 1, characterized in that, Based on the baseline LI brightness model, the estimated real-time junction temperature, estimated aging factor, and original grayscale value are used to dynamically calculate the target current to obtain a dynamically adjusted target current, including: The original grayscale values ​​are processed using a gamma correction lookup table to obtain the target brightness; The overall luminance efficiency factor is determined based on the estimated real-time junction temperature and the estimated aging factor. Divide the target brightness by the comprehensive brightness efficiency factor to obtain the compensated brightness target; The compensated brightness target is input into the reference LI brightness model to obtain the dynamically adjusted target current.

7. The LED display brightness uniformity correction method based on current feedback according to claim 6, characterized in that, The comprehensive luminance efficiency factor is determined based on the estimated real-time junction temperature and the estimated aging factor, including: using the estimated real-time junction temperature and the estimated aging factor as joint inputs, querying a preset two-dimensional joint feature model to obtain the comprehensive luminance efficiency factor.

8. The method for correcting the brightness uniformity of an LED display screen based on current feedback according to claim 7, characterized in that, Query the preset two-dimensional joint feature model, including: When the estimated real-time junction temperature and the estimated aging factor fall within the region defined by the four grid points (T1, A1), (T2, A1), (T1, A2), and (T2, A2) in the two-dimensional joint feature model, bilinear interpolation is performed to calculate the integrated luminance efficiency factor. The bilinear interpolation includes: First, linear interpolation is performed along the temperature axis to calculate the first temporary efficiency value and the second temporary efficiency value corresponding to the estimated real-time junction temperature at the first aging level A1 and the second aging level A2, respectively. Then, based on the first and second temporary efficiency values, linear interpolation is performed along the aging axis to calculate the overall luminance efficiency factor.

9. The method for correcting the brightness uniformity of an LED display screen based on current feedback according to claim 1, characterized in that, Closed-loop drive and PWM correction are performed on the original grayscale value and the dynamically adjusted target current to obtain the corrected PWM duty cycle, including: Real-time current sampling and error calculation are performed on the dynamically adjusted target current to obtain the current error signal; A baseline PWM duty cycle is generated based on the original grayscale values. The current error signal is input into the PI controller to obtain the duty cycle correction amount; The reference PWM duty cycle and the duty cycle correction amount are synthesized and saturated to obtain the corrected PWM duty cycle.

10. A brightness uniformity correction system for an LED display screen based on current feedback, characterized in that, include: The measured voltage acquisition module is used to inject low-amplitude probe current and high-amplitude probe current into the target LED pixel in response to the detection of the vertical blanking period synchronization signal to obtain the measured voltage under the low-amplitude probe current and the measured voltage under the high-amplitude probe current. The state decoupling and parameter estimation module is used to perform state decoupling and real-time parameter estimation on the measured voltage under low amplitude probe current, measured voltage under high amplitude probe current, low amplitude probe current, and high amplitude probe current based on the benchmark IVT characteristic spectrum, so as to obtain the estimated real-time junction temperature and the estimated aging factor. The raw grayscale value extraction module is used to extract the raw grayscale values ​​of the target LED pixels from the video data stream; The dynamic target current calculation module is used to calculate the target current dynamically based on the estimated real-time junction temperature, the estimated aging factor, and the original gray value using the reference LI brightness model, so as to obtain the dynamically adjusted target current. The closed-loop drive and PWM correction module is used to perform closed-loop drive and PWM correction on the original grayscale value and the dynamically adjusted target current to obtain the corrected PWM duty cycle.

Images