Intelligent power saving method for outdoor television

Abstract

The invention relates to the field of display control, and discloses an intelligent power-saving method for an outdoor television, which comprises the following steps: off-line pre-analysis: predicting an environment illumination trend in combination with a geographic position and time, and analyzing the time sequence criticality of played content; real-time perception and analysis are carried out, and instantaneous environment interference, audience presence states and spatial visual saliency of a current picture are acquired through an image acquisition unit; intelligent fusion decision is carried out, and the optimal target brightness is calculated in real time for each area of the screen through a fusion model by combining the multi-dimensional space-time information of the environment, the content, the audience and the like with pixel aging compensation data which is periodically self-calibrated and updated; and finally, converting the target brightness matrix into a hardware control instruction by the display control interface. According to the invention, non-uniform and accurate release of the power consumption in two dimensions of time and space is carried out, the total power consumption of the system can be obviously reduced, the optimal visibility of key information in any environment can be ensured, and the service life of equipment is effectively prolonged.

Description

Technical Field

[0001] This invention relates to the field of display control technology, specifically to an intelligent power-saving method for outdoor televisions. Background Technology

[0002] In recent years, with the acceleration of digitalization, outdoor television and digital signage have become increasingly widely used as important media for information dissemination and commercial advertising. These display devices are deployed in complex and variable outdoor lighting environments, and their primary technical challenge is ensuring that the screen content remains clearly visible under any weather conditions, especially in strong direct sunlight. To address this challenge, the industry common practice is to use high-brightness display panels and increase backlight power to combat ambient light interference. This strategy of directly sacrificing energy consumption for visibility, while solving the problem of display clarity to some extent, has also made outdoor televisions energy-intensive electronic devices, resulting in high operating costs and significant environmental pressure.

[0003] To alleviate this problem, some energy-saving improvement schemes have emerged in existing technologies. A common technique is to introduce one or more ambient light sensors to uniformly and macroscopically adjust the brightness of the entire screen by measuring the overall illuminance of the surrounding environment in real time. For example, the overall brightness is reduced at night or on cloudy days, and increased to maximum brightness in bright sunlight. However, this control method based on single-dimensional feedback is too crude. It treats the screen as an indivisible whole and cannot cope with the non-uniform lighting conditions commonly found in outdoor environments, such as the shadows cast by buildings or trees on one part of the screen while another part is exposed to sunlight. In this case, in order to ensure the visibility of the sunlit area, the system has to increase the brightness of the entire screen to the maximum level, resulting in unnecessary energy waste in the screen area in shadow and potentially causing visual discomfort to viewers due to excessive brightness. Furthermore, the limitations of existing technologies are also reflected in the passive and singular nature of their control logic. Their adjustment mechanism relies entirely on instantaneous and passive responses to the current ambient light, lacking the ability to proactively adapt to predictable changes in lighting (such as sunrise and sunset, and changes in the sun's trajectory caused by seasonal changes). Meanwhile, these technologies often ignore the inherent information value distribution of the displayed content, applying indiscriminate brightness output to key information areas (such as product close-ups in advertisements or headlines in news articles) and non-critical background areas. This also results in a significant amount of power being consumed in non-focus areas of visual information. Furthermore, existing systems fail to incorporate the actual presence of viewers into energy management considerations. Regardless of whether there are actual viewers in front of the screen, the device continues to operate according to preset logic, generating substantial wasted energy during periods when no one is watching.

[0004] Finally, the hardware aging problem caused by long-term high-intensity operation, especially the uneven brightness decay of the display panel's light-emitting units, is also a problem that current technology has not been able to effectively solve. Over time, areas on the screen that frequently display static images (such as logos and borders) age faster than other areas, leading to a decrease in the overall brightness and color uniformity of the screen. Existing control systems generally lack a dynamic, closed-loop self-diagnosis and compensation mechanism to combat this hardware-level degradation, resulting in a gradual deterioration of energy efficiency and display quality throughout the device's lifespan. Therefore, how to achieve an intelligent control method that can comprehensively sense the environment, content, audience, and the device's own status, and perform refined, forward-looking, and non-uniform power consumption allocation to minimize energy consumption and extend the device's lifespan while ensuring optimal visibility is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent power-saving method for outdoor televisions, aiming to solve the technical problem that existing outdoor televisions use a single, crude brightness control strategy, which makes it impossible to effectively balance high visibility and low power consumption under complex and changing actual working conditions.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent power-saving method for outdoor televisions, the method comprising: First, environmental impact information of the outdoor television screen surface is acquired. This step aims to comprehensively and accurately grasp the actual lighting environment of the screen. In a preferred embodiment, this step does not rely on a single sensor for passive measurement, but combines prediction and real-time perception modes. Specifically, the system can generate predictive environmental impact information based on a preset ambient lighting prediction model. This ambient lighting prediction model can predict the glare or shadow areas that sunlight may cause on the screen surface at different times of the day, based on the geographical location information of the outdoor television and solar trajectory data. At the same time, the system captures the actual lighting state of the screen surface in real time through an image acquisition unit (such as a camera) to generate real-time environmental interference information. This information can reflect instantaneous changes in lighting caused by unpredictable factors such as cloud changes, obstruction or reflection of surrounding objects. Finally, by dynamically fusing the predictive environmental impact information and the real-time environmental interference information, the final environmental impact information is obtained, thereby achieving a precise assessment of ambient lighting that is both predictive and responsive.

[0007] Next, the currently displayed content of the outdoor television is analyzed to obtain visual saliency information. The core of this step is to link energy-saving strategies with the efficiency of information transmission, prioritizing the allocation of limited energy to the most important display content. In a preferred embodiment, this step analyzes the content from both temporal and spatial dimensions. Temporally, the system can perform offline pre-analysis of the media content to be played, identifying key information points such as brand logos, key text, or core product images through technologies such as target detection or optical character recognition, and recording their appearance times, thereby generating temporal sequence information representing the importance distribution of content information on the timeline. Spatially, the system processes the currently displayed video frames in real time, using, for example, visual saliency detection algorithms to generate an instantaneous visual saliency map, which highlights the areas in the current image that are most attractive to the eye. Finally, by combining the temporal sequence information of content information keyness with the instantaneous visual saliency map, visual saliency information that comprehensively reflects the importance of the content is generated.

[0008] Then, based on the environmental impact information and the visual saliency information, a fusion decision is made to determine the target brightness of different areas on the outdoor television screen. This step is the core technology of the invention, which quantifies and intelligently arbitrates two heterogeneous information sources from the environment and content. In a specific embodiment, for any area on the screen... At any moment Target brightness The following model can be used for calculation:

[0009] ; in, It represents a basic brightness factor used to ensure the basic visibility of the screen; and These are the weighting coefficients for environmental impact and visual saliency, used to balance the importance of the two in decision-making; It is a normalized environmental impact value obtained based on the aforementioned environmental impact information; This is the normalized visual saliency value obtained based on the aforementioned visual saliency information. Specifically, the normalized visual saliency value... The definition can be further refined in the following ways to reflect the integration of spatiotemporal information:

[0010] ; in, The key time sequence information representing the content information is at time. The value, This indicates that the instantaneous visual saliency map is in the region. The value of .

[0011] Finally, based on the target brightness, differentiated brightness adjustments are made to the different areas. The system converts the calculated target brightness matrix into control commands for the driving units of each area of ​​the screen, achieving non-uniform and fine-grained control of the screen brightness.

[0012] Furthermore, to further improve energy efficiency and system robustness, this invention provides several preferred technical solutions. On one hand, the method may further include detecting whether there are viewers in front of the screen to generate a viewer presence factor. This viewer presence factor will be introduced into the fusion decision model. When the system detects that there are no viewers in front of the screen, it can use this factor to uniformly and deeply reduce the overall target brightness of the screen, avoiding energy waste caused by ineffective display. On the other hand, to address the problem of uneven display and decreased energy efficiency caused by LED light decay after long-term use of outdoor televisions, the method may also include an adaptive closed-loop calibration mechanism. Specifically, the system can automatically run a self-test program during a preset time period (such as late at night), controlling the screen to display a preset test pattern, while simultaneously capturing the actual response brightness of each area through the image acquisition unit. By comparing the difference between the actual response brightness and the reference brightness, the system can generate or periodically update a pixel aging compensation map. This map is then used in the fusion decision process to compensate for the target brightness of areas where light decay has occurred, thereby achieving energy savings while ensuring long-term display quality and visual uniformity.

[0013] This invention provides an intelligent power-saving method for outdoor televisions. It has the following beneficial effects:

[0014] 1. This invention achieves a shift from passive response to proactive prediction in energy saving. Unlike traditional technologies that rely solely on real-time light sensors for delayed adjustments, this invention integrates geographical location, time information, and a spherical trigonometric function model to proactively calculate the changes in the angle of incidence of sunlight on different areas of the screen throughout the day. This predictive capability allows the system to prepare in advance for impending glare or shadows by storing or reducing brightness, achieving planned energy management and avoiding drastic power fluctuations and wasted energy consumption caused by sudden changes in ambient light.

[0015] 2. This invention advances the precision of energy-saving control from macroscopic overall adjustment to microscopic spatiotemporal pixel-level precision. By simultaneously performing instantaneous visual saliency analysis and content information keyness temporal analysis, this invention can deeply understand the information value distribution of displayed content in both "space" and "time" dimensions. It no longer treats the screen as a uniformly emitting plane, but rather as a finely scalable resource, focusing its energy consumption on the climax of the advertising narrative or key products within the visuals, thereby maximizing energy efficiency while ensuring effective information delivery.

[0016] 3. This invention significantly improves the reliability of information transmission and the viewer's visual experience in complex environments. The fusion decision engine not only considers energy saving but also prioritizes ensuring visibility. It intelligently identifies areas of degraded image quality caused by real-time environmental reflections or shadows and, based on content salience, dynamically and independently compensates for the brightness of these areas. This "on-demand lighting" strategy ensures that key advertising information can still be clearly conveyed to the viewer even in corners with extremely poor lighting conditions, thereby protecting the advertiser's investment effectiveness and the brand's visual image.

[0017] 4. This invention introduces a new dimension of power consumption management based on actual viewing benefits, eliminating energy waste in unattended scenarios. By integrating audience presence factors, the system has the ability to determine whether the displayed content has an actual audience. When it detects that no one is standing in front of the screen for an extended period, the system can autonomously decide to reduce the brightness to a preset, extremely low power consumption level, rather than completely shutting it off. This design maximizes power savings during periods of inactive operation and allows for rapid screen wake-up when viewers reappear.

[0018] 5. This invention effectively slows down the hardware aging process of the device and ensures the uniformity of display quality over the long term through an adaptive closed-loop calibration mechanism. This mechanism can periodically self-diagnose the differences in light decay in different areas of the screen and generate a dynamically updated pixel aging compensation map. In subsequent brightness adjustments, the system will proactively increase the brightness of severely aged areas based on this map. This not only fundamentally solves the problem of "screen flickering" or localized dimming that may occur after long-term operation, but also extends the effective lifespan of expensive display panels through intelligent maintenance, reducing the long-term holding costs for operators. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the functional architecture of the intelligent power-saving system of the present invention; Figure 2 This is a flowchart illustrating the overall process of the intelligent power-saving method of the present invention. Figure 3 This is a schematic diagram illustrating the geometric principle of ambient lighting prediction in this invention; Figure 4 This is a schematic diagram of the data flow for multi-dimensional information fusion decision-making in this invention; Figure 5 This is a visual comparison diagram of different analytical spectra in this invention. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see the appendix Figure 1 To be continued Figure 5 This invention provides an intelligent power-saving method for outdoor televisions, which can be executed on an outdoor television terminal with specific hardware configuration and software function modules.

[0022] The hardware system of the outdoor television terminal provides the physical basis for the implementation of the method of this invention. This hardware system includes a main control unit, typically an embedded system-on-a-chip (SoC) with sufficient computing power, which integrates a central processing unit (CPU), a graphics processing unit (GPU), and a dedicated processing unit for running artificial intelligence algorithms. The hardware system also includes a display system, which consists of an outdoor LED screen panel and its supporting display driver system. The display driver system supports independent brightness control of the screen at the zone or pixel block level.

[0023] To achieve accurate perception of the external environment, the hardware system is further configured with a wide-angle high dynamic range (WDR) image acquisition unit oriented towards the screen's viewable area. The high dynamic range characteristic ensures that this unit can clearly capture details in both overly bright and overly dark areas simultaneously, even in high-contrast outdoor scenes with strong light and shadow. In addition, the hardware system also integrates a network communication unit for accessing the internet to obtain external information such as weather data, and a geolocation unit for obtaining the device's precise latitude and longitude.

[0024] The software system deployed on the main control unit is the core of the method of this invention. This software system is constructed as a series of interoperating functional modules.

[0025] The software system includes an environmental perception and prediction module. This module is responsible for processing all environmental information, and its functions cover two levels: First, it integrates an ambient lighting prediction model, which can proactively predict the lighting distribution on the screen surface over a future period based on the latitude and longitude information provided by the geolocation unit, combined with the internally stored solar position algorithm (SPA) database and real-time weather data obtained through the network communication unit; Second, it continuously processes real-time video streams from the image acquisition unit to capture unexpected, transient ambient light interference.

[0026] The software system includes a content analysis module. This module's functionality also covers two dimensions, corresponding to the temporal and spatial dimensions of the content. In the temporal dimension, when a new playlist is received, this module performs an offline pre-analysis task, calling an object detection and optical character recognition (OCR) engine to scan the media file to locate key information frames and quantify their importance. In the spatial dimension, this module runs in real-time during video playback, calling a lightweight visual saliency detection network to analyze each currently displayed frame to determine the most visually attention-grabbing regions.

[0027] The software system also includes a fusion decision engine module, which serves as the system's decision-making center. This module receives environmental information from the environmental perception and prediction module, as well as content saliency information from the content analysis module. It incorporates the core fusion decision model of this invention, responsible for quantifying and weighting this multi-dimensional and heterogeneous information, and ultimately calculating the precise target brightness value for each controllable area on the screen.

[0028] To implement the decision results, the software system includes a display control interface module. This module acts as a bridge between the upper-level software logic and the lower-level hardware driver. Its responsibility is to accurately translate the logical target brightness matrix output by the fusion decision engine module into lower-level hardware control instructions that the display driver system can recognize and execute, such as the duty cycle parameters of the pulse width modulation signal.

[0029] Finally, to ensure long-term system stability and consistent performance, the software system includes an adaptive calibration module. This module runs in the background at low frequency, responsible for scheduling and executing the screen's automated self-test program during preset idle periods (e.g., late at night). It controls the screen to display standard test patterns and, in conjunction with the image acquisition function in the environmental perception and prediction module, reverse-evaluates the luminous health status of various areas of the screen, thereby updating the internal data used for decision compensation.

[0030] Before officially entering the real-time energy-saving control cycle, the system will perform an initialization and offline pre-analysis phase to establish the necessary prior models and data foundation for subsequent intelligent decision-making. This phase mainly includes the construction and deployment of the ambient lighting prediction model, as well as in-depth time-series analysis of the content to be played.

[0031] The purpose of constructing the ambient lighting prediction model is to endow the system with the ability to anticipate major changes in ambient lighting, especially those caused by changes in the sun's position. Upon initial system deployment or location change, the geolocation unit obtains the precise longitude of the device's location. and latitude The environmental perception and prediction module uses this geographic information, combined with the current date, to... With time Using the built-in Solar Position Algorithm (SPA), the altitude angle of the sun in the sky at that moment is accurately calculated. and azimuth .

[0032] To translate the sun's position into its actual effect on the screen surface's illumination, the screen's orientation needs to be considered. Let the normal vector of the screen plane be determined by its azimuth angle. and tilt angle Defined as the angle between the sunlight and the screen's normal vector, i.e., the angle of incidence. The angle is crucial in determining the intensity of direct sunlight. This angle is calculated using spherical trigonometric functions:

[0033] in, , , All are at any time A changing function.

[0034] Subsequently, the system comprehensively considers direct sunlight, sky scattering, and weather conditions, and applies this to any area of ​​the screen. At any moment Predicted light intensity A model was created. This model integrates multiple factors:

[0035] In this model, This is a reference value for the direct solar irradiance under standard atmospheric conditions. The component of scattered light from the sky, whose intensity is related to the solar altitude angle. Related. This is a weather correction coefficient obtained from meteorological services via a network communication unit, used to quantify the attenuation effect of weather factors such as clouds on total illumination. Using this model, the system generates and stores a dynamically changing predictive glare and shadow map over time. This provides a foundational, smoothly evolving trend forecast for subsequent real-time decision-making.

[0036] Meanwhile, the content analysis module performs offline analysis of the timing criticality of the media content in the pre-arranged playlist. This process aims to pre-identify key periods of information delivery throughout the entire playback cycle. For a video file, the content analysis module utilizes its integrated object detection and OCR engine to scan the video stream frame-by-frame or by keyframe sampling.

[0037] This module has a built-in predefined "key object" library, which contains object categories whose display clarity needs to be prioritized. These include key text such as brand logos, promotional prices, and contact information in commercial advertisements, as well as core product images. During the scanning process, once the engine detects these key objects, it records their timestamp of appearance, their position in the image, their size, and their recognition confidence level.

[0038] Based on these detection results, the system generates a content information keyness time-series curve for the entire video. Any point on the curve The value is a quantitative score calculated by comprehensively considering factors such as the quantity, size, and confidence level of all key objects in the image at that moment. For example, in an image showing a large, high-confidence price tag, the corresponding... The value will be significantly higher than that of a video with only a landscape background. This curve accurately depicts the fluctuation of the "information value" of video content over time, allowing the system to predict when more power budget needs to be allocated to ensure the effective delivery of key information.

[0039] When the outdoor television enters normal playback mode, the system initiates the real-time sensing and dynamic analysis phase. This phase is executed cyclically at high frequency, and its core task is to accurately capture all dynamically changing information, including transient environmental factors, the audience's presence, and the spatial visual focus of the displayed content itself, thereby providing immediate and accurate input for the final decision.

[0040] The environmental perception and prediction module continuously processes the video stream captured by the high dynamic range image acquisition unit. In this embodiment, to accurately extract unexpected illumination changes caused by temporary obstructions or reflectors, the system first performs geometric correction and registration on the acquired image to establish a precise correspondence between its pixel coordinates and the physical area of ​​the screen. Subsequently, by applying image processing algorithms, such as high-pass filtering or calculating local standard deviation, regions in the image with drastic brightness changes are identified; these regions correspond to glare spots or sudden shadows in the real world. Through this process, the system generates a real-time environmental interference map. This map accurately depicts instantaneous environmental disturbances that cannot be covered by the prediction model.

[0041] Within the same video stream processing flow, the content analysis module performs real-time audience status detection in parallel. This function invokes a human or face detection model optimized for edge computing devices, such as a lightweight neural network, which continuously analyzes the effective viewing area in front of the screen. When an audience is detected, the system generates an audience presence signal with a value of 1. When no audience is detected in the frame, the system starts timing to record the duration of the empty state. .

[0042] Based on this detection result, the system calculates a dynamically changing audience presence factor. The calculation model for this factor is as follows:

[0043] in, It is a preset, positive constant that represents the factor value corresponding to the minimum power consumption level that the system should maintain when unattended. It is a time constant that defines the audience presence factor as it smoothly decays from 1 to approximately... The rate of brightness decay. This exponential decay model ensures that the change in brightness is a smooth transition, avoiding sudden changes in screen brightness caused by brief audience absence or detection jitter, thus improving system stability.

[0044] At the same time, for each frame currently displayed on the screen The content analysis module also performs instantaneous visual saliency analysis. This task is accomplished by a deep learning saliency detection network designed specifically for real-time applications. This network receives video frames as input, performs rapid forward propagation, and outputs a grayscale map with the same resolution as the input frames—the instantaneous visual saliency map. .

[0045] In this map, each pixel... The grayscale value directly corresponds to the intensity at which content at that location in the original image attracts human visual attention. For example, faces, moving objects, high-contrast text, or brightly colored areas in the image... The center area will appear as a bright area, while the flat background area will appear as a dark area. This map provides the system with a precise spatial distribution guide on "where the viewer is most likely to be looking at at this moment," which is an important basis for accurately allocating power to key areas of the image.

[0046] After completing multi-dimensional real-time analysis of the environment, audience, and content, the system enters the intelligent fusion decision-making and power consumption mapping stage. This stage is the central hub of the entire methodology, and its responsibility is to transform the diverse and heterogeneous information flow generated in the preceding steps into specific control commands for the brightness of various areas of the screen through a precise mathematical model.

[0047] First, the fusion decision engine module integrates and processes the various input information. It combines the predictive glare and shadow map output by the environmental perception and prediction module. and real-time environmental interference map Weighted fusion is performed to generate a final environmental impact map. This map comprehensively reflects the overall lighting environment of the screen in the current and foreseeable future.

[0048] Subsequently, to enable data from different sources to be processed within the same framework, the module normalizes key information. For any area on the screen... Its normalized environmental impact value Depend on The value at the corresponding position is linearly mapped to a specific interval. The higher the value, the stronger the light interference in that area, and the higher the demand for brightness.

[0049] Simultaneously, the module performs spatiotemporal fusion and normalization of visual saliency information. The normalized spatiotemporal content saliency value... The calculation demonstrates the deep collaboration in content understanding inherent in this invention. The calculation method is as follows:

[0050] In this calculation, The current moment is shown on the keyity time-series curve of content information obtained from the offline pre-analysis stage. The corresponding value represents the "temporal importance" of the current content within the overall playback schedule. This refers to the instantaneous visual saliency map obtained through real-time analysis in the region. The value represents the "spatial importance" of the region in the current frame. By multiplying the two values, the system can identify regions that are both at the critical information point and are the visual focus of the current frame, and assign them the highest saliency weight. The Normalize[] function is responsible for globally normalizing the result of the multiplication, ensuring that it also falls within the range of the given values. The same numerical range.

[0051] Once all inputs are ready, the fusion decision engine module invokes the core target brightness calculation model to calculate the brightness for each independently controllable area on the screen. Calculate its time at time Target brightness The model integrates all key factors:

[0052] in, It is the system's preset basic brightness factor, used to ensure that the screen has a minimum level of readability under any extreme conditions. It is a real-time detected audience presence factor used to macroscopically adjust the overall power consumption benchmark. It is obtained from the adaptive calibration module and is region-specific. The pixel aging compensation factor is used to correct brightness deviations caused by hardware aging. and These are configurable weighting coefficients used to adjust the relative priority of environmental adaptability and content clarity in the final decision, and the sum of the two is 1.

[0053] Finally, the display control interface module receives the target brightness matrix generated by the fusion decision engine module. This matrix contains the ideal brightness values ​​for all areas of the screen. The core task of the display control interface module is to convert these logical brightness values ​​into physical instructions that the underlying display driver system can execute. In this embodiment, for an LED screen that uses pulse width modulation (PWM) to control brightness, this module will convert the brightness values ​​of each area... The value is precisely converted into the duty cycle of the corresponding PWM signal through a preset mapping curve (Gamma curve). These duty cycle parameters are then sent to the LED driver chips in each area, thereby achieving non-uniform, dynamic, and highly intelligent precise control of screen power consumption.

[0054] To ensure that this method maintains the accuracy of energy-saving effects and the uniformity of the displayed image even after long-term operation of outdoor televisions, the system integrates a long-term adaptive closed-loop calibration mechanism. This mechanism is automatically activated and executed at low frequency during preset idle periods that do not affect normal business operations (e.g., early morning each day).

[0055] The calibration process is scheduled by the adaptive calibration module. Upon startup, the system first enters self-test mode. In this mode, the display control interface module bypasses normal playback content and directly instructs the display driver system to sequentially display a series of predefined standard test patterns across the entire screen. These patterns include, but are not limited to, pure white, pure red, pure green, pure blue, and uniform color fields with multiple gray levels (e.g., 50% gray). The standard luminance or chromaticity values ​​corresponding to these patterns are defined as reference luminance. .

[0056] While displaying each test pattern, the image acquisition unit simultaneously captures images. To eliminate interference from ambient light (such as streetlights at night), the system first controls the screen to a completely black state and captures a "dark field" image before displaying any test pattern. Subsequent captured test pattern images are all processed by subtracting this dark field image to obtain pure data that only reflects the screen's own luminescence. After geometric correction and registration, the system accurately measures each area on the screen from the processed image. Actual response brightness .

[0057] After obtaining the actual response brightness of each region, the adaptive calibration module begins to calculate the required aging compensation level for each region. For any region Its instantaneously calculated compensation gain factor Defined as the ratio of reference luminance to actual response luminance:

[0058] This ratio intuitively reflects the degree of light decay in the region. For example, a region whose light decays to 80% of its original brightness will have a calculated compensation gain factor of 1.25.

[0059] To avoid drastic fluctuations in calibration parameters that might result from a single measurement error, the system updates its internally stored pixel aging compensation map. At that time, a smooth update strategy was adopted. Specifically, the compensation factor ultimately stored in the map... It is obtained by fusing the newly calculated gain factor with historical values ​​using the exponential moving average algorithm:

[0060] In this formula, It is the old compensation factor value currently stored in the map. This is the latest compensation gain factor calculated during this self-test, and It is a preset update rate constant between 0 and 1. This method ensures that the evolution of the pixel aging compensation map is smooth and gradual, thereby improving the stability of the entire calibration system.

[0061] The final step in the calibration process is to apply this updated pixel aging compensation map. Persistent storage is performed. During subsequent normal playback and energy-saving control cycles, the fusion decision engine module calculates the target brightness of each area. At that time, the latest map will be used directly to obtain the compensation factor. Through this periodic closed-loop process of "self-diagnosis and self-repair," the present invention can proactively combat the aging effects of hardware, ensuring that its intelligent energy-saving strategy maintains high efficiency and accuracy throughout the entire life cycle of the equipment.

[0062] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent power saving method for an outdoor TV, characterized in that, The method comprises the following steps: obtaining environmental influence information of the outdoor TV screen surface; analyzing the display content currently displayed by the outdoor TV to obtain visual saliency information of the display content; based on the environmental influence information and the visual saliency information, making a fusion decision to determine target brightness of different regions on the outdoor TV screen; differentially adjusting the brightness of the different regions according to the target brightness.

2. The method of claim 1, wherein, The step of obtaining the environmental influence information of the outdoor TV screen surface comprises: based on a preset environmental light prediction model, generating predictive environmental influence information; through an image acquisition unit, sensing the light state of the screen surface in real time to generate real-time environmental interference information; combining the predictive environmental influence information and the real-time environmental interference information to generate the final environmental influence information.

3. The method of claim 2, wherein, The environmental light prediction model is constructed based on geographical location information and solar trajectory data of the outdoor TV, and is used to predict the direct impact of sunlight on the screen surface at different times.

4. The method of claim 1, wherein, The step of obtaining the visual saliency information of the display content comprises: offline pre-analyzing the media content to be played to generate content information key time sequence information representing the importance of the content in the time dimension; real-time processing of the currently displayed video frame to generate an instantaneous visual saliency map representing the spatial focus of the picture; combining the content information key time sequence information and the instantaneous visual saliency map to generate the final visual saliency information.

5. The method of claim 4, wherein, The offline pre-analysis comprises: identifying brand logos, key texts or core product images in the media content through target detection or optical character recognition technology to mark their positions on the time axis, thereby generating the content information key time sequence information.

6. The method of claim 1, wherein, The step of determining the target brightness of different regions on the outdoor television screen is calculated by the following model for any region At time The target brightness of the region : ； wherein, is a base luminance factor, and are weight coefficients for environmental impact and visual saliency, respectively, is a normalized environmental impact value, is a normalized visual saliency value.

7. The method of claim 6, wherein, the normalized visual saliency value is determined by ； wherein, is a value of the content information keyness temporal information at time , is a value of the instantaneous visual saliency map in region .

8. The method of claim 1, wherein, The method further comprises: detecting whether there are audiences in front of the screen through an image acquisition unit to generate an audience presence factor; combining the audience presence factor in the fusion decision, when no audience is detected, the audience presence factor is used to uniformly reduce the target brightness.

9. The method of claim 1, wherein, The method further comprises: controlling the outdoor TV to display a preset test pattern in a preset period of time, and acquiring the actual response brightness of each region of the screen through an image acquisition unit; based on the difference between the actual response brightness and the reference brightness, generating or updating a pixel aging compensation map; combining the pixel aging compensation map in the fusion decision, which is used to compensate the target brightness of the region where light decay occurs.

10. The method of claim 9, wherein, The update of the pixel aging compensation map is periodically and automatically performed to form a closed-loop calibration system for screen light decay.

Images