An adaptive image quality optimization method and device, electronic equipment and storage medium

Abstract

This application relates to the field of display processing technology, and provides a method for adaptive optimization of display images, a display terminal, and a storage medium. The method includes: acquiring the currently playing image; extracting content features from the currently playing image and performing content type recognition to obtain a content type recognition result; determining basic image quality parameters based on the content type recognition result; acquiring environmental parameters, scene parameters, content format parameters, and video parameters; generating a target image quality parameter set according to preset parameter fusion rules; performing at least one image processing step among sharpening, noise reduction, tone mapping, and image quality restoration on the currently playing image according to the target image quality parameter set, and outputting the target display image. This application can adaptively optimize the image based on different content and display conditions.

Description

Technical Field

[0001] This application relates to the field of display processing technology, and in particular to an adaptive image quality optimization method, apparatus, electronic device, and storage medium. Background Technology

[0002] Currently, digital signage, as an electronic display device that conveys information to viewers through display screens, is widely used in public places such as shopping malls, airports, train stations, and office buildings. With the development of display technology and artificial intelligence technology, image quality optimization technology for digital signage has become crucial for improving user experience and information delivery effectiveness. Existing digital signage image quality optimization technologies mainly include the following solutions: Chinese patent CN114567813A discloses an image quality improvement method, which includes parsing an input video file to determine the video frame to be played, obtaining preset parameters (including content type parameters and white point parameters) corresponding to the video frame to be played in real time, and obtaining ambient light intensity in real time through a light sensor, and dynamically adjusting the image quality of the video frame to be played according to the preset parameters and ambient light intensity. This solution can adjust the image quality according to the content type parameters and ambient light intensity, but it does not clearly define the specific classification method of content type and the corresponding image quality adjustment strategy, and lacks targeted processing for the different image quality requirements of different types of content (such as plain text, photos, videos, and charts). In addition, this solution only considers two factors: content type and ambient light intensity, without comprehensively considering other factors affecting image quality such as viewing distance, content format, and video bitrate. Chinese patent CN108810649A discloses a method for adjusting image quality, including identifying the input channel of the current source signal and obtaining the resolution and frame rate of the source signal, obtaining image quality adjustment parameters corresponding to the input channel, resolution, and frame rate from a preset mapping table, and adjusting the image signal corresponding to the source signal based on the image quality adjustment parameters. This solution selects image quality adjustment parameters based on the technical parameters (resolution, frame rate) of the source signal, but it does not perform semantic-level recognition of the image content, making it unable to distinguish between different content types such as text, photos, and videos, and difficult to perform differentiated optimization for different content image quality requirements. However, some existing technologies use cloud-based artificial intelligence image quality optimization solutions, uploading image data to a cloud server for analysis and optimization parameter calculation, and then sending the optimized parameters back to the terminal device application. The end-to-end processing latency of this solution is typically in the hundreds of milliseconds to several seconds. When the content of digital signage changes rapidly (the switching interval is usually 5 to 30 seconds), the image quality parameters cannot be adjusted in real time to keep up with the content changes, resulting in perceptible image quality jumps and affecting the user experience. Summary of the Invention

[0003] This application aims to provide an adaptive image quality optimization method, device, electronic device, and storage medium to solve the problem that existing display terminals typically use uniform image quality parameters for different types of images, or perform image quality adjustment based only on a single environmental factor, resulting in difficulty in simultaneously ensuring clarity, layering, real-time performance, and stability for different content such as text, images, videos, and charts, as well as in different deployment scenarios.

[0004] This application also aims to provide a technical solution for realizing content recognition, strategy determination and display processing link control on the display terminal side, so as to improve the real-time performance of image quality optimization processing and enhance the display terminal's adaptability to ambient light, viewing distance, dynamic range format, video bitrate and special display hardware form.

[0005] On the one hand, an adaptive image quality optimization method is provided, including: Obtain the data of the image to be played on the display terminal; The image data is preprocessed, and the preprocessed image data or image features are input into the content type recognition model in the edge artificial intelligence processing unit to obtain the content type recognition result of the current image. The target content type of the current screen is determined based on the content type recognition result; Based on the pre-stored mapping relationship between content types and image quality optimization strategies, determine the target image quality optimization strategy corresponding to the target content type; Acquire at least one piece of auxiliary decision-making information related to the current display scene, wherein the auxiliary decision-making information includes at least one of ambient light information, deployment location type information, screen dynamic range format information, and video bitrate information; Based on the target image quality optimization strategy and the at least one auxiliary decision information, determine the target display parameters of the current image and the control parameters corresponding to at least one image quality processing module; The at least one image quality processing module in the control display processing link performs image quality processing on the current screen according to the target display parameters and the control parameters, and outputs the processed screen to the display panel for display.

[0006] In one possible implementation, the target content type includes at least one of plain text, mixed text and images, photos, videos, and charts.

[0007] In one possible implementation, determining the target content type of the current frame includes: performing content type recognition on multiple consecutive frames, and statistically analyzing the recognition results of the most recent N frames, determining the majority of types in the statistical results as the target content type, where N is an integer from 3 to 5.

[0008] In one possible implementation, when the target content type is inconsistent with the target content type of the previous time step, interpolation updates are performed on the parameters corresponding to the previous image quality optimization strategy and the parameters corresponding to the current image quality optimization strategy within a preset time window to complete the smooth switching of image quality optimization strategies.

[0009] In one possible implementation, target brightness, target contrast, and target color temperature are determined based on ambient light information, and smoothing processing is performed on the ambient light information before determining the target brightness, target contrast, and target color temperature.

[0010] In one possible implementation, the deployment location type is used to characterize the viewing distance range, and at least one of the sharpening weight, text clarity enhancement weight, noise reduction weight, contrast weight, and saturation weight is adjusted according to the deployment location type.

[0011] In one possible implementation, deployment location type information is obtained through deployment location configuration issued by the management platform, and / or through target distance distribution information output by millimeter-wave radar.

[0012] In one possible implementation, when it is detected that the dynamic range format of the current image is a high dynamic range format and the display panel is a standard dynamic range display panel, the control parameters corresponding to the tone mapping module are determined so as to perform tone mapping processing on the current image.

[0013] In one possible implementation, when the current video bitrate is detected to be lower than a preset bitrate threshold, the control parameters corresponding to the bitrate repair module are determined in order to perform low bitrate repair processing on the current image.

[0014] In one possible implementation, when the resolution of the current material is lower than the physical resolution of the display panel, the control parameters corresponding to the resolution enhancement module are determined to perform super-resolution processing on the current image; when the utilization rate, temperature or power consumption of the edge AI processing unit is detected to meet the preset degradation conditions, the system switches to lightweight processing mode or pass-through display mode.

[0015] In one possible implementation, when it is detected that the display panel is an organic light-emitting diode display panel and the static elements in the current image are continuously displayed for a preset duration, the control parameters corresponding to the panel protection module are determined to control the current image to perform micro-pixel displacement processing; when it is detected that the display terminal is a splicing display terminal, the control parameters corresponding to the correction module are determined to perform brightness and color temperature correction on each sub-screen.

[0016] On the other hand, an adaptive image quality optimization device is provided, including an image acquisition unit, a recognition unit, a type determination unit, a strategy determination unit, an information acquisition unit, a parameter determination unit, and a display control unit, wherein each of the above units is used to perform the steps corresponding to the aforementioned method.

[0017] On the other hand, an electronic device is provided, including a processor, a memory, a display interface, a display processing link, and an edge artificial intelligence processing unit, wherein the memory stores program instructions, and the program instructions are executed by the processor to implement the aforementioned method.

[0018] On the other hand, a display system is provided, including a management platform and a display terminal. The management platform is used to send deployment location configuration, image quality strategy configuration, model update information and / or threshold configuration to the display terminal, and to receive distance distribution statistics and / or terminal status information reported by the display terminal; the display terminal is used to execute the aforementioned method.

[0019] On the other hand, a chip is provided, including a processing circuit and an interface circuit, wherein the processing circuit is used to invoke program instructions to execute the aforementioned method.

[0020] In another aspect, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the aforementioned method.

[0021] In another aspect, a computer program product is provided, comprising a computer program that, when run on a processor, causes an electronic device to perform the aforementioned method.

[0022] Compared with the prior art, at least one of the above technical solutions has the following beneficial effects: By performing content type recognition on the current screen and determining the corresponding image quality optimization strategy based on the identified target content type, differentiated image quality processing can be performed on different content such as text, images, videos, and charts, thereby improving the clarity and readability of the displayed content.

[0023] Since the target display parameters and the control parameters corresponding to the image quality processing module are jointly determined based on at least one of the auxiliary decision-making information, including ambient light information, deployment location type information, dynamic range format information, and video bitrate information, the adaptation accuracy of image quality parameters can be improved in different display scenarios.

[0024] Since the current image is processed by at least one image quality processing module in the display processing chain and the processed image is output to the display panel, a closed-loop control from content recognition and strategy determination to image quality processing can be formed on the terminal side, thereby improving the real-time performance of the processing.

[0025] Because it can switch to lightweight processing mode or pass-through display mode in resource-constrained scenarios, it helps to ensure playback continuity and system stability.

[0026] Because it can perform targeted processing for scenarios such as high dynamic range format playback, low bitrate video, low resolution material, OLED display panels, and splicing display terminals, it is beneficial to balance the image quality in special display scenarios with the long-term operating needs of the equipment. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the architecture of an adaptive image quality optimization system provided in an embodiment of this application; Figure 2 A flowchart illustrating content type identification and image quality strategy selection as provided in one embodiment of this application; Figure 3 A multi-factor joint decision-making flowchart is provided for one embodiment of this application; Figure 4 This is a schematic diagram of the hardware structure of a display terminal provided in an embodiment of this application; Figure 5 This is a schematic diagram of the functional unit structure of a device provided in an embodiment of this application; Figure 6 A schematic diagram of resource state-driven degradation processing provided in an embodiment of this application; Figure 7 This is a schematic diagram of organic light-emitting diode display panel protection and splicing display terminal calibration processing provided in an embodiment of this application. Detailed Implementation

[0028] The technical solution of this application will be described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the following embodiments are only used to illustrate this application and are not intended to limit the scope of protection of this application. For those skilled in the art, several substitutions and modifications can be made without departing from the technical concept of this application, and all such substitutions and modifications should fall within the scope of protection of this application.

[0029] Example 1: System Overall Architecture See Figure 1 The adaptive image quality optimization system provided in this embodiment may include a management platform 100 and a display terminal 200.

[0030] The management platform 100 can be used to distribute deployment location configurations, image quality strategy configurations, model update information, and threshold configurations to the display terminal 200. The management platform 100 can be a content management server, a terminal management platform, or a local configuration terminal. In this embodiment, the management platform 100 is not mandatory. In some scenarios, the display terminal 200 can also operate independently, completing parameter settings through local configuration.

[0031] The display terminal 200 may include: a processor 210, a memory 220, an edge artificial intelligence processing unit 230, a display processing link 240, a display interface 250, a display panel 260, an ambient light sensor 270, an optional millimeter-wave radar 280, and a communication interface 290.

[0032] The processor 210 can be used to execute control logic, call models, organize data flow, and control the collaborative work of various modules. The memory 220 can be used to store image quality optimization programs, content type recognition models, low bitrate repair models, super-resolution models, parameter tables, and configuration data. The edge-side artificial intelligence processing unit 230 can be a neural network processing unit, a graphics processor, a digital signal processor, or other dedicated processing unit supporting neural network inference. The display processing link 240 may include at least one of the following modules: scaling module, noise reduction module, sharpening module, color adjustment module, tone mapping module, bitrate repair module, resolution enhancement module, panel protection module, correction module, and output control module. The display interface 250 is used to transmit the processed image data to the display panel 260. The display panel 260 can be a liquid crystal display panel, a light-emitting diode display panel, an organic light-emitting diode display panel, or a splicing display panel. The ambient light sensor 270 is used to acquire current ambient light information. The millimeter-wave radar 280 is used to collect target distance distribution information. The communication interface 290 is used to communicate with the management platform 100, external input devices, or network content sources.

[0033] In this embodiment, after receiving the screen to be played, the display terminal 200 can identify the content type of the current screen through the edge artificial intelligence processing unit 230, and then the processor 210 combines ambient light information, deployment location type information, dynamic range format information and video bitrate information to determine the corresponding image quality optimization strategy and related control parameters, and control the display processing link 240 to output the processed screen.

[0034] The technical effect of this embodiment is that it unifies the deployment of content recognition, parameter decision-making and image quality processing on the terminal side, enabling the display terminal to have real-time, closed-loop, and adaptive image quality optimization capabilities.

[0035] Example 2: Method Flow See Figure 2 and Figure 3 This embodiment provides an adaptive image quality optimization method. This method can be executed by a display terminal or by a processor in the display terminal calling program instructions. The method may include the following steps.

[0036] Step S201: Obtain the data of the currently playing screen.

[0037] The display terminal can obtain the currently playing image data from local storage, external input interfaces, network streaming media input, or content downloaded from a content management platform. The image data can be a single frame image or the current frame from a video stream. For video streams, encoding format information, resolution information, bitrate information, and dynamic range format information can also be obtained synchronously.

[0038] Step S202: Perform content type recognition on the current screen.

[0039] The display terminal performs preprocessing on the current screen data and inputs the preprocessed screen data or screen features into the content type recognition model in the terminal artificial intelligence processing unit.

[0040] The preprocessing may include at least one of scaling, normalization, and feature extraction. The content type recognition model outputs a confidence score corresponding to at least one content type, and the processor determines the initial content type of the current frame based on the confidence scores corresponding to the at least one content type.

[0041] The content types may include one or more of the following: plain text, mixed text and images, photos, videos, and charts.

[0042] Step S203: Determine the target content type of the current screen.

[0043] To improve recognition stability, content type recognition can be performed on multiple consecutive frames, and the initial content types of the most recent N frames can be statistically analyzed. The majority of the statistical results are then determined as the target content type of the current frame, where N can be 3, 4, or 5.

[0044] In one possible implementation, when the target content type is inconsistent with the target content type of the previous moment, the display terminal does not immediately perform abrupt switching. Instead, it performs interpolation updates on the parameters corresponding to the previous image quality optimization strategy and the parameters corresponding to the current image quality optimization strategy within a preset time window to complete a smooth switching of the image quality optimization strategy, thereby reducing flickering and sudden changes in visual perception.

[0045] Step S204: Determine the target image quality optimization strategy based on the target content type.

[0046] The display terminal pre-maintains a mapping relationship between content types and image quality optimization strategies. The processor determines the target image quality optimization strategy corresponding to the target content type based on the target content type of the current screen.

[0047] For example, when the content type is plain text, strategies for enhancing text edge sharpness can be determined; when the content type is a combination of text and images, strategies for performing separate processing on text and image regions can be determined; when the content type is a photograph, strategies for noise reduction and color restoration can be determined; when the content type is video, strategies for dynamic noise reduction and smoothness enhancement can be determined; and when the content type is a chart, strategies for line enhancement and boundary enhancement can be determined.

[0048] Step S205: Obtain auxiliary decision-making information.

[0049] The display terminal acquires at least one piece of auxiliary decision-making information related to the current display scene.

[0050] The first type of auxiliary decision-making information can be ambient light information. An ambient light sensor collects the current ambient brightness, and the processor can further obtain a smoothed ambient brightness value.

[0051] The second type of auxiliary decision-making information can be deployment location type information. The deployment location type is used to characterize the viewing distance range. This information can be configured and distributed by the management platform, or it can be obtained from target distance distribution information output by millimeter-wave radar. For example, based on the statistical results of human target distance distribution within a preset time window, it can be determined whether the display terminal corresponds to a first distance range, a second distance range, or a third distance range.

[0052] The third type of auxiliary decision-making information can be the dynamic range format information of the image. If the current content is video, it can be obtained from the bitstream, container metadata, or decoding information.

[0053] The fourth type of auxiliary decision-making information can be video bitrate information. If the current content is video, it can be obtained from media information, bitrate headers, or decoding statistics.

[0054] Step S206: Determine the target display parameters and the control parameters corresponding to the image quality processing module.

[0055] The processor jointly determines the target display parameters of the current screen and the control parameters corresponding to at least one image quality processing module based on the target image quality optimization strategy and the at least one auxiliary decision information.

[0056] On the one hand, target brightness, target contrast, and target color temperature can be determined based on ambient light information. Before determining these parameters, moving average, exponential smoothing, or other smoothing processes can be applied to the ambient light information.

[0057] On the other hand, at least one of the following weights can be adjusted based on the deployment location type: sharpening weight, text clarity enhancement weight, noise reduction weight, contrast weight, and saturation weight. For example, in the deployment scenario corresponding to the first distance interval, the sharpening weight and text clarity enhancement weight are increased; in the deployment scenario corresponding to the second distance interval, the target parameter range corresponding to the sharpening weight and noise reduction weight is set; and in the deployment scenario corresponding to the third distance interval, the contrast weight and saturation weight are increased.

[0058] On the other hand, when it is detected that the dynamic range format of the current image is high dynamic range format and the display panel is a standard dynamic range display panel, the control parameters corresponding to the tone mapping module are determined in order to perform tone mapping processing.

[0059] Furthermore, when the current video bitrate is detected to be lower than the preset bitrate threshold, the control parameters corresponding to the bitrate repair module are determined in order to perform low bitrate repair processing.

[0060] Step S207: Perform image quality processing and output display.

[0061] The processor controls at least one image quality processing module in the display processing chain to perform image quality processing on the current screen according to the target display parameters and control parameters. The processed screen is then output to the display panel for display via the display interface.

[0062] In one possible implementation, when the resolution of the current material is lower than the physical resolution of the display panel, the control parameters corresponding to the resolution enhancement module can be determined to perform super-resolution processing on the current image; when the utilization rate, temperature or power consumption of the edge AI processing unit is detected to meet the preset degradation conditions, it can be switched to lightweight processing mode or pass-through display mode to ensure playback continuity.

[0063] In another possible implementation, when it is detected that the display panel is an organic light-emitting diode display panel and the static elements in the current image are continuously displayed for a preset duration, the control parameters corresponding to the panel protection module can be determined to control the current image to perform micro-pixel displacement processing; when it is detected that the display terminal is a splicing display terminal, the control parameters corresponding to the correction module can be determined to perform brightness and color temperature correction on each sub-screen.

[0064] The technical effect of this embodiment is that by combining content recognition results, auxiliary decision-making information and display processing link control, closed-loop adaptive image quality optimization is achieved on the display terminal side, and the display effect and system stability under different content types and different deployment scenarios are improved.

[0065] Example 3: Device Implementation See Figure 5This embodiment provides an adaptive image quality optimization device. This device can be implemented as a software functional module, or as a hardware circuit, firmware logic, or a combination of hardware and software. The device may include: The image acquisition unit is used to acquire the image data to be played on the display terminal. The recognition unit is used to preprocess the image data and input the preprocessed image data or image features into the content type recognition model in the edge artificial intelligence processing unit to obtain the content type recognition result of the current image. A type determination unit is used to determine the target content type of the current screen based on the content type identification result. The strategy determination unit is used to determine the target image quality optimization strategy corresponding to the target content type based on the pre-stored mapping relationship between content types and image quality optimization strategies. An information acquisition unit is used to acquire at least one piece of auxiliary decision-making information related to the current display scene; The parameter determination unit is used to determine the target display parameters of the current screen and the control parameters corresponding to at least one image quality processing module based on the target image quality optimization strategy and the at least one auxiliary decision information. The display control unit is used to control at least one image quality processing module in the display processing link to perform image quality processing on the current screen according to the target display parameters and the control parameters, and output the processed screen to the display panel for display.

[0066] In one possible implementation, the recognition unit may include a preprocessing subunit, a model invocation subunit, and an initial determination subunit. The preprocessing subunit is used to perform at least one of scaling, normalization, and feature extraction on the image; the model invocation subunit is used to input the preprocessed image data or image features into the content type recognition model; and the initial determination subunit is used to determine the initial content type based on the confidence level corresponding to at least one content type output by the model.

[0067] In one possible implementation, the type determination unit can also be used to statistically analyze the initial content types of multiple consecutive frames and determine the majority of the types in the statistical results as the target content type.

[0068] In one possible implementation, the strategy determination unit may include a mapping management subunit and a switching control subunit. The mapping management subunit is used to maintain a mapping table between content types and image quality optimization strategies; the switching control subunit is used to perform parameter interpolation and smooth switching when the content type changes.

[0069] In one possible implementation, the information acquisition unit may include an ambient light acquisition subunit, a deployment location acquisition subunit, a format parsing subunit, and a bitrate detection subunit. The ambient light acquisition subunit is used to acquire ambient light information; the deployment location acquisition subunit is used to receive deployment location configurations issued by the management platform and / or receive target distance distribution information output by millimeter-wave radar; the format parsing subunit is used to acquire the dynamic range format information of the current image; and the bitrate detection subunit is used to acquire the bitrate information of the current video.

[0070] In one possible implementation, the device may further include a format processing unit, a bitrate repair unit, a resolution enhancement unit, a panel protection unit, and a correction unit. The format processing unit performs tone mapping processing on the current image when it detects that the dynamic range format of the current image is a high dynamic range format and the display panel is a standard dynamic range display panel. The bitrate repair unit performs low bitrate repair processing on the current image when it detects that the current video bitrate is lower than a preset bitrate threshold. The resolution enhancement unit performs super-resolution processing on the current image when the resolution of the current source material is lower than the physical resolution of the display panel. The panel protection unit performs micro-pixel displacement processing on the current image when it detects that the display panel is an organic light-emitting diode display panel and static elements in the current image have been continuously displayed for a preset duration. The correction unit performs brightness and color temperature correction on each sub-screen when it detects that the display terminal is a splicing display terminal.

[0071] The technical advantage of this embodiment is that it adopts the form of functional units to support each step in the method embodiment and clarifies the input, processing and output relationships of each unit, thereby providing clearer structural and implementation support for the device claims.

[0072] Example 4: Implementation by an electronic device See Figure 4 This embodiment provides an electronic device. This electronic device can be a digital signage unit, a commercial display terminal, a smart interactive flat panel, an advertising machine, a splicing control terminal, or other devices with display capabilities.

[0073] The electronic device includes a processor, memory, display interface, display processing link, edge AI processing unit, display panel, and communication interface. The processor executes control logic, calls models, organizes data flow, and controls the collaborative work of various modules. The memory stores image quality optimization programs, content type recognition models, low bitrate repair models, super-resolution models, parameter tables, and configuration data. The edge AI processing unit executes content type recognition model inference. The display processing link performs at least one of the following processing operations on the image: scaling, noise reduction, sharpening, color adjustment, tone mapping, and output control. The display interface outputs the processed image data to the display panel. The communication interface receives configuration and content data from the management platform.

[0074] The memory stores program instructions, which, when executed by the processor, cause the electronic device to perform the following operations: Get the data of the currently playing frame; The edge-side AI processing unit is invoked to perform content type recognition on the current screen. The target content type is determined based on the content type identification results, and the target image quality optimization strategy is determined based on the mapping relationship between the content type and the image quality optimization strategy. Acquire at least one of the following: ambient light information, deployment location type information, dynamic range format information, and video bitrate information; Determine the target display parameters and control parameters corresponding to at least one image quality processing module based on the target image quality optimization strategy and the at least one auxiliary decision information; At least one image quality processing module in the control display processing link performs image quality processing on the current screen and outputs it to the display panel via the display interface.

[0075] In one possible implementation, the edge AI processing unit can be integrated into the system-on-a-chip (SoC) or exist as a standalone coprocessor. The display processing chain can be either a dedicated hardware module or the processor calling an image processing program within a unified memory architecture.

[0076] The technical advantage of this embodiment is that it provides a clear hardware implementation path for the method, making it easy to cover the entire product.

[0077] Example 5: System Collaboration Implementation This embodiment provides a display system, including a management platform and a display terminal.

[0078] The management platform is used for unified configuration and management of multiple display terminals. It can distribute deployment location configurations, image quality strategy configurations, model versions, threshold parameters, resolution enhancement switches, panel protection switches, and calibration task parameters to different display terminals. Upon receiving the corresponding configurations, the display terminals can write them into their local parameter tables and participate in subsequent multi-factor joint decision-making.

[0079] In one implementation, the display terminal is equipped with millimeter-wave radar. The display terminal can determine the deployment location type based on statistical results of human target distance distribution within a preset time window, and report these statistical results to a management platform. The management platform can confirm or manually override the automatically identified deployment location type based on the information reported by the display terminal, and issue an updated deployment location configuration to the display terminal.

[0080] For example, in a chain store scenario, the management platform can configure the checkout terminal for close-range deployment in the first distance interval, the shelf end terminals for mid-range deployment in the second distance interval, and the entrance corridor terminals for long-range deployment in the third distance interval. Different display terminals can then apply different sharpening, noise reduction, contrast, and saturation weights to the same type of content accordingly.

[0081] The technical effect of this embodiment is that, without changing the main real-time processing link on the terminal side, the collaboration between the management platform and the display terminal improves deployment flexibility, unified parameter management capabilities, and operation and maintenance efficiency in multi-terminal scenarios.

[0082] Example 6: Special Display Hardware Extension Processing 1. Protective treatment for organic light-emitting diode display panels See Figure 7 When the display panel is an organic light-emitting diode (OLED) display panel, the processor can detect static elements in the current screen. These static elements can be logos, borders, subtitle areas, or fixed menu areas that remain in place for extended periods.

[0083] When a static element is detected to be continuously displayed for a preset duration, the processor can determine the control parameters corresponding to the panel protection module and control the current screen to perform micro-pixel displacement processing. The displacement can be performed in at least one direction, either horizontal or vertical, and the displacement amplitude can be set to a pixel-level offset less than the user's visually significant perception threshold. This method reduces the risk of image retention caused by prolonged high-load operation of local pixels.

[0084] 2. Splicing display terminal calibration processing When the display terminal is a splicing display terminal, the brightness and color temperature parameters of each sub-screen can be collected separately, and the parameter compensation amount of each sub-screen can be obtained according to the calibration model. Subsequently, the processor can determine the control parameters corresponding to the calibration module, and send the target brightness parameters and target color temperature parameters to the display control module of each sub-screen through the calibration unit, so that each sub-screen has a high degree of consistency when splicing display.

[0085] The technical effect of this embodiment is that it further extends the image quality optimization link to panel protection and multi-screen consistency correction, thereby improving the long-term use effect under special hardware form factors.

[0086] Example 7: Resource State-Driven Degradation Processing See Figure 6 In this embodiment, the display terminal can perform resource status awareness and priority scheduling for multiple tasks on the edge AI processing unit. For example, content type recognition tasks and image quality optimization tasks can be set as high-priority tasks, while model updates, log analysis, or other non-real-time auxiliary tasks can be set as low-priority tasks.

[0087] When the utilization rate of the edge AI processing unit exceeds a preset threshold, the chip temperature exceeds a preset threshold, or the system enters a power-constrained state, the processor can prioritize the execution of high-priority tasks while reducing the scheduling frequency of low-priority tasks, pausing some low-priority tasks, or switching at least one of the resolution enhancement processing and low bitrate repair processing to a lightweight processing mode. Under more stringent degradation conditions, it can also switch to pass-through display mode. In this way, the stable operation of critical image quality processing links can be guaranteed in resource-constrained scenarios, and the system robustness can be improved.

[0088] The technical advantage of this embodiment is that it prioritizes the execution of critical image quality processing tasks in resource-constrained scenarios, while also taking into account the system's continuous operation capability and display stability.

[0089] It should be noted that the above embodiments can be implemented individually or in combination. Technical features disclosed in one embodiment can also be applied to other embodiments without conflict.

[0090] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by hardware related to program instructions. The program can be stored in a computer-readable storage medium. When executed, the program can implement the steps in the above embodiments. The computer-readable storage medium may include a read-only memory, random access memory, a magnetic disk, an optical disk, flash memory, or other non-transitory readable storage media.

[0091] Furthermore, the device units in the above embodiments can be implemented in software, hardware, or a combination of both.

Claims

1. An adaptive image quality optimization method, characterized in that, include: Obtain the data of the image to be played on the display terminal; The image data is preprocessed, and the preprocessed image data or image features are input into the content type recognition model in the edge artificial intelligence processing unit to obtain the content type recognition result of the current image. The target content type of the current screen is determined based on the content type recognition result. Based on the pre-stored mapping relationship between content types and image quality optimization strategies, determine the target image quality optimization strategy corresponding to the target content type; Acquire at least one piece of auxiliary decision-making information related to the current display scene, wherein the auxiliary decision-making information includes at least one of ambient light information, deployment location type information, screen dynamic range format information, and video bitrate information; Based on the target image quality optimization strategy and the at least one auxiliary decision information, determine the target display parameters of the current image and the control parameters corresponding to at least one image quality processing module; The at least one image quality processing module in the control display processing link performs image quality processing on the current screen according to the target display parameters and the control parameters, and outputs the processed screen to the display panel for display.

2. The method according to claim 1, characterized in that, The target content type includes at least one of plain text, mixed text and images, photos, videos, and charts.

3. The method according to claim 1, characterized in that, The process of preprocessing the image data and inputting the preprocessed image data or image features into the content type recognition model in the edge AI processing unit to obtain the content type recognition result of the current image includes: Perform at least one of the following preprocessing steps on the image data: scaling, normalization, and feature extraction; The preprocessed image data or image features are input into the content type recognition model to obtain the confidence level corresponding to at least one content type. The initial content type of the current screen is determined based on the confidence level corresponding to the at least one content type.

4. The method according to claim 1 or 3, characterized in that, Determining the target content type of the current screen based on the content type recognition result includes: Perform content type recognition on multiple consecutive frames to obtain the initial content type corresponding to the multiple consecutive frames; Statistical analysis of the initial content types of the most recent N frames; The majority of the types in the statistical results are identified as the target content type; Where N is the content type; Where N is an integer from 3 to 5.

5. The method according to claim 4, characterized in that, When it is determined that the target content type is inconsistent with the target content type of the previous time, within a preset time window, interpolation update is performed based on the parameters corresponding to the previous image quality optimization strategy and the parameters corresponding to the current image quality optimization strategy to complete the smooth switching of image quality optimization strategy.

6. The method according to claim 1, characterized in that, Determining the target display parameters of the current screen based on the target image quality optimization strategy and at least one piece of auxiliary decision information includes: Determine the target brightness, target contrast, and target color temperature based on ambient light information; When the ambient light intensity increases, the target brightness and contrast are increased, and the target color temperature is adjusted to the first color temperature range; When the ambient light intensity decreases, reduce the target brightness and adjust the target color temperature to the second color temperature range; Before determining the target brightness, target contrast, and target color temperature, the ambient light information is smoothed.

7. The method according to claim 1, characterized in that, The deployment location type is used to characterize the viewing distance range; The step of determining the target display parameters of the current screen and the control parameters corresponding to at least one image quality processing module based on the target image quality optimization strategy and the at least one piece of auxiliary decision information includes: When the deployment location type corresponds to the first distance interval, increase the sharpening weight and text clarity enhancement weight; When the deployment location type corresponds to the second distance interval, set the target parameter range corresponding to the sharpening weight and the noise reduction weight; When the deployment location type corresponds to the third distance interval, increase the contrast weight and saturation weight.

8. The method according to claim 7, characterized in that, The deployment location type information is obtained through at least one of the following methods: Receive deployment location configurations from the management platform; The system receives target distance distribution information from millimeter-wave radar and determines the deployment location type based on the target distance distribution statistics within a preset time window.

9. The method according to claim 1, characterized in that, Determining the control parameters corresponding to at least one image quality processing module based on the at least one piece of auxiliary decision information includes at least one of the following: When it is detected that the dynamic range format of the current image is high dynamic range format and the display panel is a standard dynamic range display panel, the control parameters corresponding to the tone mapping module are determined in order to perform tone mapping processing on the current image. When the current video bitrate is detected to be lower than the preset bitrate threshold, the control parameters corresponding to the bitrate repair module are determined in order to perform low bitrate repair processing on the current image.

10. The method according to claim 1, characterized in that, The step of determining the control parameters corresponding to at least one image quality processing module based on the at least one piece of auxiliary decision information further includes: When the resolution of the current material is lower than the physical resolution of the display panel, determine the control parameters corresponding to the resolution enhancement module in order to perform super-resolution processing on the current image; When the utilization rate, temperature, or power consumption of the edge AI processing unit is detected to meet the preset degradation conditions, the system switches to lightweight processing mode or pass-through display mode.

11. The method according to claim 1, characterized in that, The step of determining the control parameters corresponding to at least one image quality processing module based on the at least one piece of auxiliary decision information further includes: When it is detected that the display panel is an organic light-emitting diode display panel and the static elements in the current screen are continuously displayed for a preset time, the control parameters corresponding to the panel protection module are determined to control the current screen to perform micro-pixel displacement processing. When the display terminal is detected to be a splicing display terminal, the control parameters corresponding to the calibration module are determined in order to perform brightness and color temperature calibration on each sub-screen.

12. An adaptive image quality optimization device, characterized in that, include: The image acquisition unit is used to acquire the image data to be played on the display terminal. The recognition unit is used to preprocess the image data and input the preprocessed image data or image features into the content type recognition model in the edge artificial intelligence processing unit to obtain the content type recognition result of the current image. A type determination unit is used to determine the target content type of the current screen based on the content type identification result. The strategy determination unit is used to determine the target image quality optimization strategy corresponding to the target content type based on the pre-stored mapping relationship between content types and image quality optimization strategies. An information acquisition unit is used to acquire at least one piece of auxiliary decision-making information related to the current display scene, wherein the auxiliary decision-making information includes at least one of ambient light information, deployment location type information, screen dynamic range format information, and video bitrate information; The parameter determination unit is used to determine the target display parameters of the current screen and the control parameters corresponding to at least one image quality processing module based on the target image quality optimization strategy and the at least one auxiliary decision information. The display control unit is used to control at least one image quality processing module in the display processing link to perform image quality processing on the current screen according to the target display parameters and the control parameters, and output the processed screen to the display panel for display.

13. The apparatus according to claim 12, characterized in that, The type determination unit is used for: Perform content type recognition on multiple consecutive frames to obtain the initial content type corresponding to the multiple consecutive frames; The initial content types of the most recent N frames are statistically analyzed, and the majority of the types in the statistical results are identified as the target content type. Where N is an integer from 3 to 5; The display control unit is also used to perform interpolation updates within a preset time window based on the parameters corresponding to the previous image quality optimization strategy and the parameters corresponding to the current image quality optimization strategy when the target content type changes, so as to complete the smooth switching of image quality optimization strategies.

14. The apparatus according to claim 12, characterized in that, The information acquisition unit includes an ambient light acquisition subunit, used to acquire ambient light information; The parameter determination unit is used to perform smoothing processing on the ambient light information before determining the target brightness, target contrast and target color temperature, and to determine the target brightness, target contrast and target color temperature based on the smoothed ambient light information.

15. The apparatus according to claim 12, characterized in that, The information acquisition unit includes a deployment location acquisition subunit, used to receive deployment location configurations issued by the management platform and / or receive target distance distribution information output by millimeter-wave radar; The parameter determination unit is used to adjust at least one of the sharpening weight, text clarity enhancement weight, noise reduction weight, contrast weight, and saturation weight according to the deployment location type.

16. The apparatus according to claim 12, characterized in that, Also includes: The format processing unit is used to perform tone mapping processing on the current image when it is detected that the dynamic range format of the current image is a high dynamic range format and the display panel is a standard dynamic range display panel. The bitrate repair unit is used to perform low bitrate repair processing on the current frame when the current video bitrate is detected to be lower than a preset bitrate threshold. The resolution enhancement unit is used to perform super-resolution processing on the current image when the resolution of the current material is lower than the physical resolution of the display panel. The panel protection unit is used to perform micro-pixel displacement processing on the current screen when it is detected that the display panel is an organic light-emitting diode display panel and the static elements in the current screen have been continuously displayed for a preset time. The calibration unit is used to perform brightness and color temperature calibration on each sub-screen when the display terminal is detected to be a splicing display terminal.

17. An electronic device, characterized in that, The device includes a processor, a memory, a display interface, a display processing link, and an edge-side artificial intelligence processing unit. The memory stores program instructions, which, when executed by the processor, cause the electronic device to perform the method of any one of claims 1 to 11.

18. A display system, characterized in that, This includes the management platform and display terminal; The management platform is used to send deployment location configuration, image quality strategy configuration, model update information and / or threshold configuration to the display terminal, and to receive distance distribution statistics and / or terminal status information reported by the display terminal. The display terminal is used to perform the method according to any one of claims 1 to 11.

19. A chip, characterized in that, It includes a processing circuit and an interface circuit, wherein the processing circuit is used to invoke program instructions to execute the method of any one of claims 1 to 11.

20. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method of any one of claims 1 to 11.

21. A computer program product, characterized in that, It includes a computer program that, when run on a processor, causes an electronic device to perform the method of any one of claims 1 to 11.

Images