Power consumption monitoring and self-optimizing system of energy-saving high-definition LED screen

Abstract

The invention relates to the technical field of LED control optimization, in particular to a power consumption monitoring and self-optimizing system of an energy-saving high-definition LED screen. The method comprises the following steps: monitoring various data in a monitoring module; further acquiring the power extended range of the target frame picture in the power consumption analysis module and acquiring the motion energy density of each type of picture element; further according to the distribution change of the lamp beads activated between the adjacent frames by each type of picture elements, in combination with the motion energy density, obtaining the picture motion intensity of the target frame picture; acquiring a real-time extra power consumption component in an adjusting module; constructing a database of impression coefficients; and finally, based on the real-time impression coefficient, in combination with the additional power consumption component and the distribution of the mass center of each lamp bead relative to the current thermodynamic diagram, the brightness of the lamp beads is adjusted, the problem that energy consumption redundancy and impression decline of the LED screen coexist is effectively solved, and the balance of energy conservation and visual experience is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of LED control optimization, and in particular to a power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen. Background Art

[0002] With the rapid development of LED display technology, high-definition LED screens are widely used in outdoor advertising, public information displays, and other applications. However, their high energy consumption is becoming increasingly prominent. Current energy-saving technologies primarily achieve partial savings through adaptive ambient light control, power supply optimization, and driver chip improvements. For example, ambient light sensors can monitor and adjust screen brightness in real time, reducing inefficient power consumption under strong daylight conditions.

[0003] However, traditional methods lack the balance between energy saving and image quality. Their energy efficiency is unstable in complex scenarios (such as the coexistence of dynamic images and static areas), and they are prone to over-dimming or contrast distortion. Indiscriminately reducing brightness or turning off regional backlighting can also cause color cast and halo effects, significantly reducing the user experience. Summary of the Invention

[0004] In order to solve the technical problem that traditional energy-saving adjustment methods are difficult to balance user experience and energy-saving effects, the purpose of the present invention is to provide a power consumption monitoring and self-optimization system for energy-saving high-definition LED screens. The technical solutions adopted are as follows:

[0005] A power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen, the system comprising:

[0006] Monitoring module: obtains the current heat map and its centroid, current data and ambient brightness of each frame of the LED screen, and captures various screen elements; selects any frame as the target frame;

[0007] Power consumption analysis module: obtains the power range of the target frame image based on the change of the centroid of the current heat map of adjacent frames; analyzes the change of pixels of each type of the image element in adjacent frames based on the optical flow method to obtain the motion energy density of each type of image element in the target frame image; obtains the image motion intensity of the target frame image based on the distribution change of the activated lamp beads of each type of image element in adjacent frames and the motion energy density;

[0008] Adjustment module: According to the difference in the following mode of the electric range extender according to the picture motion intensity of the real-time picture and the historical picture, combined with the real-time current, obtain the real-time additional power consumption component; according to the picture performance of different ambient brightness and the picture motion intensity, obtain the perception coefficient and establish a test database; based on the real-time picture motion intensity and real-time ambient brightness, extract the perception coefficient of the test database, and adjust the brightness of the lamp beads in combination with the additional power consumption component and the distribution of the center of mass of each lamp bead relative to the current heat map.

[0009] Furthermore, the method for obtaining the motion energy density includes:

[0010] The motion speed of each pixel between adjacent frames is measured based on the optical flow method. The ratio of the current value of a single pixel when it is fully lit to the square of the maximum brightness value is used as the equivalent energy consumption to convert pixel motion into pixels.

[0011] The motion energy density of each type of picture element is obtained according to the motion speed of all pixels of each type of picture element, the equivalent energy consumption and the preset switching brightness state factor.

[0012] Furthermore, the method for obtaining the picture motion intensity includes:

[0013] Select the screen elements as target elements one by one; obtain the activation complexity of the target element in each row based on the number of lamp beads required to be activated in each row when the target element appears in the target frame, combined with the discrete degree of brightness change of the lamp beads in this row in adjacent frames; use the same method to obtain the activation complexity of the target element in each column;

[0014] Obtaining an activation power consumption ratio of the target element in the target frame according to a proportional feature of the activation complexity of the target element in the row and column with the same serial number;

[0015] The motion intensity of the target frame is obtained by fusing the motion energy density with the activation power consumption ratio based on each type of the picture elements.

[0016] Furthermore, the method for obtaining the additional power consumption component includes:

[0017] Performing time-series alignment on the curves of the motion intensity of the image and the electric range extension, obtaining a historical following error based on the degree of dispersion of the difference between the motion intensity and the electric range extension in all history; and using the difference between the real-time motion intensity and the electric range extension as the real-time following error;

[0018] According to the difference between the real-time following error and the historical following error, combined with the average value of the electric range extension curve and the real-time total screen current, the real-time additional power consumption component is obtained.

[0019] Furthermore, the method for obtaining the perception coefficient includes:

[0020] Based on the preset ambient brightness and the image motion intensity, the brightness of the LED screen is reduced from high to low within the brightness range of the ambient brightness, and a binocular stereo vision camera is used to obtain the display image of the LED screen;

[0021] Obtaining the contrast degradation rate based on the average brightness change of the screen area before and after the brightness reduction, combined with the brightness standard deviation of the screen before the brightness reduction;

[0022] The halo visibility index is obtained based on the brightness gradient change at the edge of the screen area before and after the brightness is reduced, combined with the distance from the edge point to the center of the screen;

[0023] Obtaining a blur coefficient according to a change in the ratio of low-frequency component energy to high-frequency component energy in the picture area of ​​the screen before and after the brightness is reduced;

[0024] Based on the contrast degradation rate, halo visibility index, and blur coefficient, the perception coefficient under the conditions of the ambient brightness and the image motion intensity during the test is obtained.

[0025] Furthermore, the method for adjusting the brightness of the lamp beads includes:

[0026] Allocate the additional power consumption component based on the distance of each lamp bead from the centroid of the real-time current thermodynamic map and the real-time perception coefficient to obtain the maximum allowable power reduction of each lamp bead;

[0027] The lamp current is adjusted within the maximum allowable reduced power range to adjust the lamp brightness.

[0028] Furthermore, when adjusting the brightness of the lamp beads, a bilateral filter can be used to balance power saving and edge sharpness to smoothly adjust the brightness of the lamp beads.

[0029] Furthermore, the method for obtaining the electric range extension includes:

[0030] The rate of change of the position of the center of mass of the current thermodynamic diagram between the target frame and the adjacent previous frame is used as the electric power range of the target frame.

[0031] Furthermore, the method for obtaining the centroid of the current thermodynamic map includes:

[0032] The LED screen is mapped to a coordinate system, and the ratio of the current value of each lamp bead to the total current value of the entire screen is used to perform weighted summation on the horizontal coordinate and vertical coordinate of each lamp bead to obtain the horizontal coordinate and vertical coordinate of the center of mass.

[0033] Furthermore, the method for capturing various screen elements includes:

[0034] The real-time display screen is captured by an HDMI2.1 capture card, and various screen elements in the screen are segmented using semantics; the screen elements include at least text, background, and dynamic objects.

[0035] The present invention has the following beneficial effects:

[0036] The present invention first obtains various data through a monitoring module to provide a basis for subsequent analysis. It further obtains the electric range extension based on the change in the centroid of the current heat map between adjacent frames, characterizing the additional energy consumption caused by the migration of the power consumption distribution. It further analyzes the change in the pixels of each type of screen element between adjacent frames to obtain the motion energy density and quantify the motion energy intensity consumed by the pixels moving between frames. It further obtains the picture motion intensity based on the distribution change of the lamp beads activated by each type of screen element between adjacent frames, combined with the motion energy density, and accurately evaluates the intensity of screen drive resource occupation under different motion trends, providing a quantitative basis for subsequent power consumption optimization. It further reflects the consistency deviation of the screen drive power consumption control based on the difference in the picture motion intensity displayed by the electric range extension following mode in the real-time picture. In combination with the real-time current, it can be used to obtain the additional ineffective power consumed in addition to catering to the picture changes, providing a basis for adaptive optimization. It further establishes a database of the perception coefficient corresponding to the picture performance of different ambient brightness and picture motion intensity to provide a basis for ensuring the user's viewing experience. Finally, based on the real-time perception coefficient, combined with the additional power consumption component and the distribution of each lamp bead relative to the centroid of the current heat map, the lamp bead brightness is adjusted. This solution monitors screen motion, current heat maps, and ambient brightness to identify excess power consumption and visual perception coefficients. Combined with the thermal distribution of the lamp beads, it dynamically adjusts the brightness, effectively resolving the coexistence of redundant energy consumption and reduced visual perception on LED screens, and achieving a balance between energy saving and visual experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 A system block diagram of a power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen provided by one embodiment of the present invention;

[0039] Figure 2 A physical diagram of an LED screen provided by one embodiment of the present invention;

[0040] Figure 3A schematic diagram of an LED module provided by one embodiment of the present invention;

[0041] Figure 4 A schematic diagram comparing the image motion intensity and electric range extension before alignment provided by one embodiment of the present invention;

[0042] Figure 5 A schematic diagram showing a comparison between aligned image motion intensity and electric range extension provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of a power consumption monitoring and self-optimization system for an energy-saving, high-definition LED screen proposed in accordance with the present invention. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0044] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0045] The following describes in detail a specific scheme of a power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen provided by the present invention in conjunction with the accompanying drawings.

[0046] See also Figure 1 , which shows a system block diagram of a power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen provided by an embodiment of the present invention. The system includes: a monitoring module 101, a power consumption analysis module 102 and an adjustment module 103.

[0047] In one embodiment of the present invention, in response to the technical problem that the current mainstream power-saving strategy for high-definition LED screens sacrifices non-core content to save power, which seriously affects the normal function and picture quality of the screen, this application monitors the additional consumption during dynamic image switching and executes a power-saving strategy on the LED screen within a reasonable range. This can restrain the impact of indiscriminate power saving on picture quality fluctuations, ensure user experience, reduce additional power consumption as much as possible, and improve resource utilization.

[0048] Monitoring module 101: obtains the current heat map and its centroid, current data and ambient brightness of each frame of the LED screen, and captures various screen elements; selects any frame as the target frame.

[0049] The core of the LED screen is the LED module, which is the display core and consists of LED lamp beads, driver chips, PCB circuit boards, etc. Figure 2 , which shows a physical diagram of an LED screen provided by an embodiment of the present invention; the packaging form of the LED screen is divided into indoor unit board (such as 8×8 dot matrix) and outdoor module box.

[0050] See also Figure 3 , which shows a schematic diagram of an LED module provided by an embodiment of the present invention, Figure 3 The figure shows an 8×8 lamp bead matrix unit board, which contains 16 pins, divided into 8 rows and 8 columns, and each lamp bead corresponds to one pixel.

[0051] In one embodiment of the present invention, the activation process adopts a row-by-row scanning method, cyclically activating 8 rows at intervals of 1ms-5ms. When the Nth row is scanned, a high level is output to the row pin (anode), and the column pin (cathode) is synchronously controlled to output a low level, lighting the target lamp bead;

[0052] The scanning cycle is controlled within 16.67ms (corresponding to a 60Hz refresh rate), and the residual characteristics of the human eye are used to form a stable image.

[0053] The magnetoresistive current sensor is integrated into the LED driver IC, and each sensor covers an 8×8 pixel unit, and then the I 2 The C bus is aggregated to the central processor to generate a full-screen current thermal map (the current thermal map and the real-time image need to be aligned in time and space), and its resolution matches the sensor density.

[0054] It should be noted that the current heat map does not show the current at each lamp bead position, but the current at each sensor coverage area. The current at each lamp bead position is only an approximation and cannot be accurately read.

[0055] By monitoring the real-time current at each row and column pin position, accurate current data passing through each lamp bead can be obtained;

[0056] Real-time measurement of the total current, instantaneous voltage and screen temperature data of the entire LED screen;

[0057] Use an HDMI2.1 capture card to capture the real-time display image, and use semantic segmentation to separate the text, background, dynamic objects and other units in the image; dynamic objects can include scrolling text, icons, etc.

[0058] Environmental sensors, including light sensors and temperature sensors, are used to monitor ambient brightness and temperature.

[0059] Considering that the centroid of the current heat map marks the center of the power consumption distribution, in order to facilitate the subsequent analysis of the additional power consumption caused by power migration during screen switching, the centroid of the current heat map of each frame of the LED screen is obtained at the same time.

[0060] Preferably, in one embodiment of the present invention, the LED screen is mapped to a coordinate system, and the ratio of the current value of each lamp bead to the total current value of the entire screen is weightedly summed up on the horizontal and vertical coordinates of each lamp bead to obtain the horizontal and vertical coordinates of the center of mass.

[0061] Among them, when the horizontal coordinate and the vertical coordinate of each lamp bead are weighted and summed respectively, the weighted weight is the ratio of the current value of each lamp bead to the total current value of the entire screen.

[0062] It should be noted that the analysis method for each frame is consistent. Here, any frame is selected as the target frame and described using the target frame as an example. In actual scenarios, real-time frames can be analyzed one by one as the target frame over time.

[0063] It should be noted that the working principles of LED screens and various sensors, as well as semantic segmentation technology, are already well known to those skilled in the art and will not be elaborated here.

[0064] Power consumption analysis module 102: obtains the power range of the target frame image based on the change of the center of mass of the current heat map of adjacent frames; analyzes the change of pixels of each type of image element in adjacent frames based on the optical flow method, and obtains the motion energy density of each type of image element in the target frame image; obtains the image motion intensity of the target frame image based on the distribution change of the lamp beads activated by each type of image element between adjacent frames and the motion energy density.

[0065] Considering that the motion of adjacent frames may cause the power consumption distribution to shift, which may lead to additional power consumption, the power range extension of the target frame is obtained according to the change of the centroid of the current heat map of adjacent frames.

[0066] Preferably, in one embodiment of the present invention, the rate of change of the position of the centroid of the current thermodynamic diagram of the target frame and the adjacent previous frame is used as the electric power range of the target frame.

[0067] Among them, the distance between the center of mass of the current thermodynamic diagram of the target frame and the adjacent previous frame is taken as the change in the center of mass position, and the time interval between the target frame and the adjacent previous frame is taken as the time change. Therefore, based on the ratio of the center of mass position change and the time change as the electric range extension, the rate of change of the center of mass of the current thermodynamic diagram of the adjacent frames is characterized.

[0068] Since different elements in the picture transform, different types of pixel motion consume different amounts of energy. Considering that the optical flow method can accurately capture the motion trend at the pixel level and match the pixels in adjacent frames to analyze the energy consumption characteristics of pixel motion, the optical flow method is used to analyze the changes in pixels of each type of picture element in adjacent frames, obtain the motion energy density of each type of picture element in the target frame, and quantify the motion energy intensity consumed by the pixel motion between frames, reflecting the dynamic picture quality fidelity cost.

[0069] Preferably, in one embodiment of the present invention, considering that the greater the pixel movement speed, the more frequent the current changes and the higher the energy consumption, and in LED screen displays, the luminous brightness of the pixel is usually in a linear relationship with the driving current with a quadratic exponential, the movement speed of each pixel between adjacent frames is measured based on the optical flow method; the ratio of the current value of a single pixel when it is fully lit to the square of the maximum brightness value is used as the equivalent energy consumption of the pixel to convert the pixel movement into the pixel, and the energy consumption of the unit brightness movement is estimated, which effectively reflects the instantaneous power consumption change caused by the pixel brightness change, and facilitates the construction of a dynamic image quality optimization mechanism for energy consumption perception;

[0070] Considering that the lamp beads have two actions, namely maintaining brightness and switching brightness, the motion energy density is analyzed here, so the energy consumption factor of switching brightness also needs to be referred to;

[0071] Based on this, the motion energy density of each type of picture element is obtained according to the motion speed, equivalent energy consumption and preset switching brightness state factor of all pixels of each type of picture element.

[0072] As an example, the target frame is compared with the adjacent previous frame; the resources consumed in maintaining the brightness of the lamp beads and switching the brightness state are set to half each, and the preset switching brightness state factor is 0.5. The square of the movement speed of each pixel, the equivalent energy consumption and the preset switching brightness state factor are multiplied as the motion energy density of each pixel; the average motion energy density of all pixels of a type of picture element in the target frame is taken as the motion energy density of this type of picture element.

[0073] It should be noted that the current value when a single pixel is fully lit can be measured in advance; the energy consumption ratio of the two actions of the lamp bead, namely maintaining brightness and switching brightness, is affected by factors such as design and environment, and the implementer can adjust it according to actual conditions; the optical flow method is a technical means well known to those skilled in the art.

[0074] Because the row driver tube requires a higher voltage, more frequent switching operations, and greater gate capacitance charging and discharging power during the row switching process, the power loss of the row driver tube is higher than that of the column driver tube. Therefore, for the screen with vertical migration (row switching during vertical migration), the power loss will be higher.

[0075] And because the motion energy density of each type of screen element is different, the number and distribution of row drives activated by different screen elements at the same motion distance are also different, which leads to significant differences in their impact on the overall power consumption. Therefore, according to the distribution changes of the lamp beads activated by each type of screen element between adjacent frames, combined with the motion energy density, the screen motion intensity of the target frame is obtained, so that the system can combine the distribution of screen elements with the actual driving path to accurately evaluate the occupancy intensity of screen driving resources under different motion trends, providing a quantitative basis for subsequent power consumption optimization.

[0076] Preferably, in one embodiment of the present invention, screen elements are selected as target elements by category, and the screen elements are analyzed one by one;

[0077] Considering that the more lamps that need to be activated in each row, the more complex the activation control. At the same time, the more discrete the brightness changes of the lamps in adjacent frames, the more complex the brightness changes, and the more complex the activation control. Based on this, the activation complexity of the target element in each row is obtained according to the number of lamps that need to be activated in each row when the target element in the target frame appears in this row, combined with the discrete degree of the brightness changes of the lamps in this row in adjacent frames.

[0078] As an example, adjacent frames refer to the target frame and the adjacent previous frame. The brightness change is the difference between the brightness of each lamp bead in the selected row in the target frame and its brightness in the adjacent previous frame. The discrete degree of the brightness change is expressed by the variance. When the target element appears in each row, the product of the number of lamp beads required to be activated in this row and the variance of the brightness change of the lamp beads in this row in the adjacent frame pictures is used as the activation complexity of the target element in each row.

[0079] The same method is used to obtain the activation complexity of the target element in each column. The product of the number of lamp beads required to be activated in each column when the target element appears and the variance of the brightness change of the lamp beads in this column in adjacent frames is used as the activation complexity of the target element in each column.

[0080] Considering that the ratio of "row activation complexity / column activation complexity" reflects to a certain extent the imbalance of energy consumption caused by the position distribution of the element, it represents the distribution change of the lamp beads activated by each type of screen element in adjacent frames. The larger the ratio, the more significant its driving energy consumption;

[0081] It further integrates the motion energy density that characterizes the impact of content changes on power consumption, and the activation power consumption ratio that characterizes the impact of position distribution on power consumption, to comprehensively reflect the animation load and more accurately reflect the actual motion intensity and hardware power consumption pressure of the target frame, providing support for dynamic optimization.

[0082] Based on this, the activation power consumption ratio of the target element in the target frame is obtained according to the proportional characteristics of the activation complexity of the target element in the rows and columns with the same sequence number;

[0083] The motion intensity of the target frame is obtained by fusing the motion energy density based on the activation power consumption ratio of each type of screen element.

[0084] As an example, the row and column numbers refer to Figure 3 , the average value of the ratio of the activation complexity of the target element in the row and column with the same sequence number is used as the activation power consumption ratio of the target element in the target frame.

[0085] The activation complexity of the row corresponds to the numerator position in the calculation ratio, and the activation complexity of the column corresponds to the denominator position. To prevent the denominator from being zero, the activation complexity of the column can be preprocessed by dividing it by zero, such as adding a constant 1 to the activation complexity of the column and using it as the denominator.

[0086] The activation power consumption ratio of each type of picture element in the target frame is used as the numerator, the sum of the activation power consumption of all types of picture elements in the target frame is used as the denominator, and the fractional ratio is used as the weighted summation weight. The motion energy density is weightedly summed with the obtained weighted summation weight, and the result is used as the picture motion intensity of the target frame.

[0087] Adjustment module 103: According to the difference in the follow-up mode of the electric range extender based on the picture motion intensity of the real-time picture and the historical picture, combined with the real-time current, obtain the real-time additional power consumption component; according to the picture performance of different ambient brightness and picture motion intensity, obtain the perception coefficient and establish a test database; based on the real-time picture motion intensity and real-time ambient brightness, extract the perception coefficient of the test database, and adjust the brightness of the lamp beads in combination with the additional power consumption component and the distribution of the center of mass of each lamp bead relative to the current heat map.

[0088] Taking into account that the pin current responds first and the lamp bead brightness changes later, the change in the electric range extension is earlier than the picture movement. The picture motion intensity has a certain following mode for the electric range extension, and the difference in the following mode between the real-time picture and the historical picture reflects the consistency deviation of the picture drive power consumption control. Combined with the real-time current, it can reflect the additional invalid power consumed in addition to catering to the picture changes. Therefore, according to the difference in the following mode of the picture motion intensity of the real-time picture and the historical picture for the electric range extension, combined with the real-time current, the real-time additional power consumption component is obtained.

[0089] Preferably, in one embodiment of the present invention, the method for obtaining the real-time excess power consumption component includes:

[0090] After linear normalization of the image motion intensity and power range in the time series to remove the dimension in their respective data dimensions, the least squares fitting is performed on the time series. Figure 4 , which shows a schematic diagram comparing the image motion intensity and electric range extension before alignment provided by an embodiment of the present invention. Figure 4 Corresponding to a progressive opening animation displayed on the LED screen, the screen elements increase as time goes by, text and icons are displayed, and the intensity of the screen movement increases. Then the screen gradually dims, the opening animation ends, new elements appear, and the intensity of the screen movement increases again. Figure 4 It can be seen that the change in electric range extension occurs a little earlier than the movement in the picture.

[0091] Align the time series of the curves of the image motion intensity and the power range to eliminate the response delay. Figure 5 , which shows a schematic diagram comparing the aligned picture motion intensity and the electric range extension provided by an embodiment of the present invention, wherein the alignment is performed using a cross-correlation method, specifically: calculating the cross-correlation function between the two signals; finding the peak of the cross-correlation function, and the time offset corresponding to the peak is the time delay between the two signals; and performing time shifting on one of the signals according to the calculated time delay to align with the other signal.

[0092] Considering that the motion intensity and electric range extension of the LED screen show certain regularity under long-term statistics, the historical following error is obtained based on the discrete degree of the difference between all historical motion intensity and electric range extension, which can be used to establish an error tolerance range; the difference between the real-time motion intensity and electric range extension is used as the real-time following error;

[0093] Taking into account that the greater the electric range extension, the higher the power consumption level, in order to reduce the error and misjudgment caused by the fluctuation of the overall power level of the screen, when quantifying the difference between the real-time following error and the historical following error, it is also necessary to combine the average of the electric range extension curve. Therefore, based on the difference between the real-time following error and the historical following error, combined with the average of the electric range extension curve and the real-time total current of the screen, the real-time additional power consumption component is obtained.

[0094] As an example, the difference between the real-time following error and the historical following error is used as the numerator, the mean of the electric range extension curve is used as the denominator, and the fractional ratio is linearly normalized to quantify the difference in the following mode of the electric range extension due to the motion intensity of the real-time image and the historical image. Finally, the product of the normalized result and the total current of the real-time screen is used as the real-time additional power consumption component, providing a basis for subsequent adaptive optimization of power consumption control.

[0095] It should be noted that regular alignment corrections, such as once every minute, are set to prevent further time delays after alignment. During the continuous self-optimization analysis process, the analysis time domain of historical data can be limited, such as to the last two days, to reduce the amount of computation. Linear normalization is normalization within the corresponding data dimension. This method can be used for normalization in the embodiments of the present invention. The specific technical means are well known to those skilled in the art and will not be elaborated here.

[0096] Taking into account that when the brightness of the lamp beads is reduced to save energy, the visual perception of different picture intensities and different ambient brightnesses is different. In order to save energy while ensuring the user viewing experience, it is also necessary to analyze the user perception. Therefore, based on the picture performance of different ambient brightness and picture motion intensity, the perception coefficient is obtained and a test database is established.

[0097] Preferably, in one embodiment of the present invention, the test is performed at a frontal viewing angle, without considering viewing angle differences; and only in weather with good sunlight, without considering reflections, extreme weather conditions, and other random factors; the minimum-maximum brightness range of the external environment is obtained by an environmental sensor, and the brightness range is evenly divided into 10 parts, with the highest brightness of each part being the brightness of a brightness level; a pre-prepared animation containing all screen motion intensities is played, and played once before and after the brightness is reduced, and the same time sequence frames are compared to compare and analyze the detailed features of the screen image before and after the brightness reduction;

[0098] The brightness level of the environment was changed, and a pre-prepared animation was played repeatedly. Each brightness level and each intensity of motion in the animation was used as a preset experimental condition. The brightness of the LED screen was reduced from high to low within the brightness range of the ambient brightness at the preset ambient brightness and motion intensity. A binocular stereo vision camera was used to capture the display image of the LED screen. The correlation coefficients of the three types of perception, contrast, halo, and blur, were analyzed after the brightness of the LED screen was reduced. In each experiment, the screen brightness of the LED screen was the maximum ambient brightness and 70% of the maximum ambient brightness. The analysis process was consistent under each experimental condition and is described here as just one example.

[0099] Considering that the human eye's perception of picture details mainly comes from picture contrast, which is manifested in the brightness difference of the picture, the brightness standard deviation represents the degree of picture detail, and the average brightness change before and after brightness reduction reflects the absolute degree of brightness attenuation. Therefore, the contrast degradation rate is obtained based on the average brightness change of the picture area of ​​the screen before and after brightness reduction, combined with the brightness standard deviation of the picture before brightness reduction.

[0100] As an example, the difference between the average brightness of the screen area before brightness reduction and the average brightness of the screen area after brightness reduction is used as the numerator, the brightness standard deviation of the screen before brightness reduction is used as the denominator, and the fractional ratio is used as the contrast degradation rate.

[0101] When the brightness of LED screens is reduced, a halo effect is prone to occur in the edge areas of the image. In particular, the halo near the center has a greater impact on the human eye. The greater the brightness gradient change, the more obvious the edge halo. Therefore, the halo visibility index is obtained based on the brightness gradient change at the edge of the screen area before and after the brightness is reduced, combined with the distance from the edge point to the center of the picture.

[0102] As an example, the unit length of the distance is one pixel, and the distance from the edge point to the center of the picture is used as the independent variable, and the softmax[0.1 x ] function is used as the weighted weight to perform weighted summation on the absolute value of the brightness gradient difference at the edge of the screen area before and after the brightness is reduced, and the result is used as the halo visibility index.

[0103] Among them, softmax[] is the normalization function, x is the independent variable, and the brightness gradient change decays exponentially with increasing distance.

[0104] Considering that the essence of blur is the loss of high-frequency details of the image, the more the frequency domain energy is concentrated in the low frequency, the blurrier the image is. Therefore, the blur coefficient is obtained based on the change in the ratio of the low-frequency component energy to the high-frequency component energy in the image area of ​​the screen before and after the brightness is reduced.

[0105] As an example, the screen image before and after the brightness is reduced is converted into the frequency domain, and the high and low frequencies are divided using low-pass filtering. The ratio of the total energy of each low-frequency component to the total energy of the high-frequency component is calculated respectively, and the difference between the ratio of the frequency components corresponding to the brightness before the brightness is reduced and the ratio of the frequency components corresponding to the brightness after the brightness is reduced is used as the fuzzy coefficient.

[0106] Based on the contrast degradation rate, halo visibility index, and blur coefficient, the perception coefficient under the conditions of ambient brightness and picture motion intensity during the test is obtained.

[0107] As an example, the linear normalization results of the Euclidean norm of the contrast degradation rate, halo visibility index, and blur coefficient are used as the perception coefficients under the conditions of ambient brightness and picture motion intensity during the corresponding test. Finally, the picture performance of ambient brightness and picture motion intensity is quantified from three perception angles: contrast degradation rate, halo visibility index, and blur coefficient. The smaller the perception coefficient, the smaller the impact of brightness reduction on perception.

[0108] It should be noted that in other embodiments of the present invention, the implementer can divide the ambient brightness levels into more detailed levels, adjust the way the LED screen descends, and the pre-prepared animation can be selected by the implementer. Frame animations containing the same motion intensity can only compare and analyze one frame to save resources. The more types of motion intensity included in the picture, the more detailed the division of the ambient brightness levels, the more experimental resources are consumed, and the more detailed the visual perception coefficient is, the richer the scenes corresponding to the data in the test database.

[0109] It should be noted that Fourier transform can be used to extract the frequency domain information of the image, and the total energy of the low- and high-frequency components is the sum of the squares of the amplitude spectrum in the low-frequency region and the sum of the squares of the amplitude spectrum in the high-frequency region, respectively; the low-pass filter can adopt an IIR low-pass filter (R can be set to 0.5) or other low-pass filters such as FIR low-pass filter, which are technical means well known to those skilled in the art and will not be repeated here.

[0110] The perception coefficient can be used as an indicator of picture quality perceived by the human eye, reflecting the degree of negative impact of reducing brightness on visual quality under the current environment and screen motion intensity conditions; the additional power consumption component serves as a redundant energy consumption indicator, reflecting the energy consumption space for compressing brightness; the distribution of lamp beads relative to the centroid of the current heat map reflects the importance of different lamp beads in the current picture. Therefore, the perception coefficient of the test database is extracted based on the real-time picture motion intensity and real-time ambient brightness. Combined with the additional power consumption component and the distribution of the centroid of each lamp bead relative to the current heat map, the brightness of the lamp beads is adjusted, which can dynamically identify and compress redundant energy consumption while ensuring the viewing quality.

[0111] Preferably, in one embodiment of the present invention, considering that the smaller the perception coefficient, the smaller the impact of reducing the brightness on the perception, and at the same time, the larger the additional power consumption component, the larger the energy consumption space that can be compressed, and the larger the allowable power reduction; and considering that the farther the lamp bead is from the centroid of the real-time current heat map, the more likely it is a non-core pixel, and the larger the allowable space for reducing the brightness, the distance between the lamp bead and the centroid of the real-time current heat map represents the distribution of the lamp bead relative to the centroid of the current heat map;

[0112] Based on this, the extra power consumption component is allocated according to the distance of each lamp bead from the centroid of the real-time current heat map and the real-time perception coefficient to obtain the maximum allowable power reduction for each lamp bead;

[0113] Adjust the lamp current and brightness within the maximum allowable power reduction range.

[0114] As an example, based on the real-time picture motion intensity and the real-time ambient brightness, the corresponding perception coefficient under the closest test conditions (the absolute value of the difference is the smallest) is matched in the test database; the product of 1 minus the difference in the perception coefficient, the additional power consumption component and the distance of each lamp bead from the center of mass of the current thermogram is used as the numerator, the sum of the distances of all lamp beads from the center of mass of the current thermogram is used as the denominator, and the fractional ratio is used as the maximum allowable reduction in power for each lamp bead.

[0115] Finally, within the maximum allowable range of power reduction, the appropriate dynamic control amount is selected; brightness gradient constraints can be implemented on adjacent lamp bead groups (the gradient threshold of the brightness of adjacent lamp beads after adjustment is set), or a bilateral filter can be used to balance power saving and edge sharpness to obtain a smoother lamp bead brightness adjustment amount; all of these are well-known technologies and will not be elaborated on.

[0116] In summary, in response to the technical problem that traditional energy-saving adjustment methods are difficult to balance user experience and energy-saving effects, the present invention proposes a power consumption monitoring and self-optimization system for energy-saving high-definition LED screens. The present invention first monitors various data in the monitoring module; further obtains the power range of the target frame image and the motion energy density of each type of image element in the target frame image in the power consumption analysis module; further obtains the image motion intensity of the target frame image based on the distribution changes of the lamp beads activated by each type of image element between adjacent frames, combined with the motion energy density; in the adjustment module, according to the difference in the follow-up mode of the image motion intensity of the real-time image and the historical image to the power range, combined with the real-time current, obtains the real-time additional power consumption component; constructs a database of perception coefficients; finally, based on the real-time perception coefficient, combined with the additional power consumption component and the distribution of the center of mass of each lamp bead relative to the current heat map, adjusts the brightness of the lamp beads. This solution identifies the additional power consumption and perception coefficient by monitoring the image motion, current heat map and ambient brightness, and dynamically adjusts the brightness in combination with the heat distribution of the lamp beads, effectively solving the problem of the coexistence of energy consumption redundancy and perception degradation of the LED screen, and achieving a balance between energy saving and visual experience.

[0117] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0118] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A power consumption monitoring and self-optimization system for energy-saving high-definition LED screens, characterized in that: The system comprises: Monitoring module: obtains the current heat map and its centroid, current data and ambient brightness of each frame of the LED screen, and captures various screen elements; selects any frame as the target frame; Power consumption analysis module: obtains the power range of the target frame image based on the change of the centroid of the current heat map of adjacent frames; analyzes the change of pixels of each type of the image element in adjacent frames based on the optical flow method to obtain the motion energy density of each type of image element in the target frame image; obtains the image motion intensity of the target frame image based on the distribution change of the activated lamp beads of each type of image element in adjacent frames and the motion energy density; Adjustment module: According to the difference in the following mode of the electric range extender according to the picture motion intensity of the real-time picture and the historical picture, combined with the real-time current, obtain the real-time additional power consumption component; according to the picture performance of different ambient brightness and the picture motion intensity, obtain the perception coefficient and establish a test database; based on the real-time picture motion intensity and real-time ambient brightness, extract the perception coefficient of the test database, and adjust the brightness of the lamp beads in combination with the additional power consumption component and the distribution of the center of mass of each lamp bead relative to the current heat map.

2. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 1 is characterized in that: The method for obtaining the motion energy density includes: The motion speed of each pixel between adjacent frames is measured based on the optical flow method. The ratio of the current value of a single pixel when it is fully lit to the square of the maximum brightness value is used as the equivalent energy consumption to convert pixel motion into pixels. The motion energy density of each type of picture element is obtained according to the motion speed of all pixels of each type of picture element, the equivalent energy consumption and the preset switching brightness state factor.

3. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 1 is characterized in that: The method for obtaining the picture motion intensity includes: Select the screen elements as target elements one by one; obtain the activation complexity of the target element in each row based on the number of lamp beads required to be activated in each row when the target element appears in the target frame, combined with the discrete degree of brightness change of the lamp beads in this row in adjacent frames; use the same method to obtain the activation complexity of the target element in each column; Obtaining an activation power consumption ratio of the target element in the target frame according to a proportional feature of the activation complexity of the target element in the row and column with the same serial number; The motion intensity of the target frame is obtained by fusing the motion energy density with the activation power consumption ratio based on each type of the picture elements.

4. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 1 is characterized in that: The method for obtaining the additional power consumption component includes: Performing time-series alignment on the curves of the motion intensity of the image and the electric range extension, obtaining a historical following error based on the degree of dispersion of the difference between the motion intensity and the electric range extension in all history; and using the difference between the real-time motion intensity and the electric range extension as the real-time following error; According to the difference between the real-time following error and the historical following error, combined with the average value of the electric range extension curve and the real-time total screen current, the real-time additional power consumption component is obtained.

5. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 1 is characterized in that: The method for obtaining the perception coefficient includes: Based on the preset ambient brightness and the image motion intensity, the brightness of the LED screen is reduced from high to low within the brightness range of the ambient brightness, and a binocular stereo vision camera is used to obtain the display image of the LED screen; Obtaining the contrast degradation rate based on the average brightness change of the screen area before and after the brightness reduction, combined with the brightness standard deviation of the screen before the brightness reduction; The halo visibility index is obtained based on the brightness gradient change at the edge of the screen area before and after the brightness is reduced, combined with the distance from the edge point to the center of the screen; Obtaining a blur coefficient according to a change in the ratio of low-frequency component energy to high-frequency component energy in the picture area of ​​the screen before and after the brightness is reduced; Based on the contrast degradation rate, halo visibility index, and blur coefficient, the perception coefficient under the conditions of the ambient brightness and the image motion intensity during the test is obtained.

6. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 1, characterized in that: The method for adjusting the brightness of the lamp beads includes: Allocate the additional power consumption component based on the distance of each lamp bead from the centroid of the real-time current thermodynamic map and the real-time perception coefficient to obtain the maximum allowable power reduction of each lamp bead; The lamp current is adjusted within the maximum allowable reduced power range to adjust the lamp brightness.

7. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 6, characterized in that: When adjusting the brightness of the lamp beads, a bilateral filter can also be used to balance power saving and edge sharpness to smoothly adjust the brightness of the lamp beads.

8. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 1 is characterized in that: The method for obtaining electric range extension includes: The rate of change of the position of the center of mass of the current thermodynamic diagram between the target frame and the adjacent previous frame is used as the electric power range of the target frame.

9. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 1, characterized in that: The method for obtaining the centroid of the current thermodynamic map includes: The LED screen is mapped to a coordinate system, and the ratio of the current value of each lamp bead to the total current value of the entire screen is used to perform weighted summation on the horizontal coordinate and vertical coordinate of each lamp bead to obtain the horizontal coordinate and vertical coordinate of the center of mass.

10. The power consumption monitoring and self-optimization system for an energy-saving high-definition LED screen according to claim 1, characterized in that: The method for capturing various screen elements includes: The real-time display screen is captured by an HDMI2.1 capture card, and various screen elements in the screen are segmented using semantics; the screen elements include at least text, background, and dynamic objects.

Images