CN-122002077-A - Display screen data processing method and system

Abstract

The invention provides a display screen data processing method and system, wherein the method comprises the steps of S1, obtaining original picture data to be displayed to form a source display data signal, S2, carrying out real-time analysis and semantic segmentation on the source display data signal, generating a content characteristic signal containing each region attribute label according to the source display data signal, S3, synchronously monitoring the current running load, available transmission bandwidth and power consumption state of a data processing system, generating a system state monitoring signal, S4, inputting the content characteristic signal and the system state monitoring signal into a strategy decision engine together, wherein the strategy control signal comprises compression coding strategies, rendering priorities and transmission scheduling instructions assigned to different content regions, and S5, generating and outputting a driving display signal for driving a target display screen to display pictures according to the strategy control signal. The invention solves the problems of low efficiency and high energy consumption caused by the fact that the display screen cannot intelligently allocate resources when processing the mixed content.

Inventors

• YAO DAZHI
• HUANG YONGZHEN
• WU ZHAOFEI

Assignees

• 河南环宸新能源科技有限公司

Dates

Publication Date

20260508

Application Date

20260123

Claims (10)

1. 1. A display screen data processing method, comprising: S1, acquiring original picture data to be displayed to form a source display data signal; S2, carrying out real-time analysis and semantic segmentation on the source display data signals, identifying a static image area, a dynamic video area and an interactive text area in the source display data signals, and generating content characteristic signals containing attribute labels of all the areas according to the static image area, the dynamic video area and the interactive text area; s3, synchronously monitoring the current running load, the available transmission bandwidth and the power consumption state of the data processing system, and generating a system state monitoring signal; S4, inputting the content characteristic signals and the system state monitoring signals to a strategy decision engine, wherein the strategy decision engine performs fusion analysis on the signals according to a preset optimizing rule set so as to output strategy control signals for dynamically configuring a data processing flow, and the strategy control signals comprise compression coding strategies, rendering priorities and transmission scheduling instructions assigned to different content areas; And S5, according to the strategy control signal, adaptively scheduling and calling a corresponding hardware encoder, a graphic rendering pipeline and a data transmission channel, performing differentiated compression, rendering and packaging processing on different areas in the source display data signal, and generating and outputting a driving display signal for driving a target display screen to display pictures.
2. 2. The method for processing display screen data according to claim 1, wherein the real-time analysis and semantic segmentation of the source display data signals specifically means that through a special analysis module deployed at the front end of a display processing pipeline, continuous frame data is subjected to frame-by-frame pixel semantic understanding and region division, the process firstly utilizes a lightweight neural network model to perform feature extraction on data of a frame buffer, identifies edge, texture, motion vector and color distribution features in an image, and then segments a picture into a plurality of rectangular or non-rectangular region blocks with different semantic attributes based on the features, wherein the attributes at least comprise but are not limited to a completely static background region, a dynamic video region containing periodic or non-periodic motion and an interactive text region containing characters or graphic user interface elements, each identified region is endowed with a unique attribute tag, and the tag and coordinate information and motion state descriptors of the region together form the content feature signal to provide refined content context information for subsequent policy decisions.
3. 3. The method for processing display screen data according to claim 1, wherein the current operation load, available transmission bandwidth and power consumption state of the synchronous monitoring data processing system are a multi-parameter parallel acquisition and fusion process, the operation load monitoring is realized by inquiring a performance counter and a hardware performance unit provided by an operating system kernel, the core utilization rate, the memory bandwidth occupancy rate and the depth of a frame processing queue of a central processing unit and a graphic processing unit are obtained in real time, the available transmission bandwidth monitoring is evaluated for the residual data throughput capacity of an internal bus bandwidth and an external display interface of the system and is calculated according to an interface protocol standard and real-time data flow, the power consumption state monitoring is realized by reading current, voltage and power readings provided by a power management unit and comprehensively judging the energy consumption level and the heat dissipation state of the system by combining thermal sensor data of equipment, and all the acquired original parameters are fused and packaged into a structured system state monitoring signal after normalization and time stamp alignment processing.
4. 4. The method of claim 1, wherein the policy decision engine is provided with a configurable optimization rule set and a lightweight evaluation model, wherein the optimization rule set is composed of a plurality of condition-action rules, each rule defines an optimal processing policy to be adopted under the combination of a specific content characteristic signal and a system state monitoring signal, the evaluation model is used for arbitrating and selecting according to a preset optimization target when a plurality of applicable rules conflict or a plurality of policy options exist, the optimization target comprises but is not limited to minimization of overall power consumption, maximization of visual quality fidelity of a specific content area or minimization of system end-to-end processing delay, after the engine receives the content characteristic signal and the system state monitoring signal, feature matching and rule triggering are firstly performed, then a final policy is calculated through the evaluation model, the finally generated policy control signal is a composite instruction comprising a plurality of sub instruction sets, each sub instruction set precisely corresponds to an identified content area, and the configuration parameters of a follow-current path of the area data to be processed in the follow-current path are specified.
5. 5. The method of claim 1, wherein the compression coding strategy included in the strategy control signal is dynamically assigned according to the attribute and system state of the content area, a lossless or visual lossless compression algorithm is assigned to the signal portion identified as the still image area to preserve details when the system resources are abundant, a high-compression lossy algorithm is assigned to save bandwidth and computational power when the system resources are intense, a dynamic video area is adaptively selected and parameter optimized among multiple video coding schemes such as inter-frame predictive coding, intra-frame compression and variable code rate coding according to the motion complexity and the available bandwidth of the system, a specific coding or rendering enhancement strategy capable of ensuring the sharpness and definition of the character edges is always assigned to the interactive text area to prevent text blurring, and the dynamic assignment process ensures that the data processing is always kept in a self-adaptive balance between visual quality and processing efficiency under complex mixed content scenes and variable system environments.
6. 6. The method for processing display screen data according to claim 1, wherein the policy control signal includes a rendering priority instruction for scheduling rendering resources of a graphics processor, the instruction allocates appropriate rendering queue priorities and computing resource quotas to different content areas according to importance degrees of areas in the content feature signal and load conditions of the graphics processor in the system state monitoring signal, a high priority area, such as a focus window or a key animation element where a user is interacting, is allocated with more shader core processing time and earlier rendering time to ensure smoothness, and a low priority background or inactive area, the rendering instruction of which may be appropriately delayed or processed by using a simplified shader program, the priority-based dynamic rendering scheduling mechanism effectively avoids interaction jams caused by average allocation of rendering resources, and improves perception speed and user experience of system response.
7. 7. The method according to claim 1, wherein the adaptively scheduling and invoking the corresponding hardware encoder means that the policy execution module dynamically selects one or more available dedicated encoding hardware units in the system according to the instruction of the policy control signal and configures the working parameters thereof, when the processing requirement is mainly video encoding with high throughput, the dedicated video encoding hardware unit is scheduled, when the processing requirement is mainly still image encoding with high fidelity, another set of image processing units which are more good for such algorithm may be scheduled, meanwhile, the system can determine whether to perform parallel encoding or serial encoding according to the load condition, and even process different areas with different hardware encoders in one frame of picture, the scheduling process ensures that the encoding task is always allocated to the current most suitable hardware unit with the highest energy efficiency ratio for execution, thereby maximizing the utilization of the hardware advantage of the heterogeneous computing platform, improving the overall encoding efficiency and reducing the burden of the main processor.
8. 8. The method of claim 1, wherein the differential packaging process performed on different areas in the source display data signal according to the policy control signal is a final integration step performed after all area data is compressed and rendered, the step allocates a transmission identifier, a priority label and address information of a target buffer area to each area data packet according to a transmission scheduling instruction in the policy control signal, the data packets of the high-priority real-time interaction area are marked as urgent and are sent on the most preferable transmission time slot and the high-speed data channel, the low-priority background update data can be packaged into larger data blocks and transmitted by using the idle bandwidth or transmitted in a buffer delay, and the differential data packaging and scheduling mechanism enables the limited display interface bandwidth to be utilized most effectively, and the updating speed of the most critical picture part for user experience is guaranteed preferentially, so that unnecessary transmission congestion and delay are reduced.
9. 9. The method of claim 1 further comprising a closed-loop feedback adjustment mechanism, wherein after the driving display signal is output to the target display screen and a frame of image is displayed, the system collects actual processing result feedback signals, the feedback signals include actual processing time consumption, final output code rate and actual power consumption increment information sensed by the system sensor, the feedback signals are sent back to the policy decision engine to be compared and analyzed with input signals and decision expectations of the current frame of policy, the policy decision engine utilizes the comparison result to perform fine adjustment and learning on an optimization rule set or evaluation model parameter in the policy decision engine, so that next decision for similar scenes can be closer to actual system performance and optimization targets, and the closed-loop feedback mechanism enables the whole data processing system to have self-adaptive learning and continuous optimization capability and to be self-adjusted along with changes of usage patterns and environments.
10. 10. A system using a display screen data processing method according to any one of claims 1 to 9, comprising: The data module acquires original picture data to be displayed and forms a source display data signal; The analysis module is used for carrying out real-time analysis and semantic segmentation on the source display data signals, identifying a static image area, a dynamic video area and an interactive text area in the source display data signals, and generating content characteristic signals containing attribute labels of all the areas according to the static image area, the dynamic video area and the interactive text area; The detection module synchronously monitors the current running load, the available transmission bandwidth and the power consumption state of the data processing system and generates a system state monitoring signal; The optimizing module inputs the content characteristic signals and the system state monitoring signals to a strategy decision engine, and the engine performs fusion analysis on the signals according to a preset optimizing rule set so as to output strategy control signals for dynamically configuring a data processing flow, wherein the strategy control signals comprise compression coding strategies, rendering priorities and transmission scheduling instructions assigned to different content areas; And the display module is used for adaptively scheduling and calling corresponding hardware encoders, graphics rendering pipelines and data transmission channels according to the strategy control signals, performing differentiated compression, rendering and packaging processing on different areas in the source display data signals, and generating and outputting driving display signals for driving a target display screen to display pictures.

Description

Display screen data processing method and system Technical Field The invention relates to the technical field of optimization and self-adaptive control of display screen data processing, in particular to a display screen data processing method and system. Background With the rapid development of display screen technology to high resolution, high refresh rate and high dynamic range, the amount of display data that needs to be generated, processed and transmitted per unit time increases exponentially, which constitutes an unprecedented challenge for the bandwidth, computing power and system power consumption of the entire data processing link. In conventional display processing architecture, data processing typically employs a fixed or preset pipeline, such as uniformly applying a compression algorithm to the entire frame, rendering at a fixed quality, and transmitting at a constant priority. This "one-shot" mode of processing is imperative in the face of increasingly complex modern application scenarios. Modern user interfaces and content are often mixtures of still images, dynamic videos, interactive text and graphic elements, and content of different nature has great differences in the requirements of processing fidelity, real-time. For example, text information requires extremely high edge definition to ensure readability, dynamic video requires a steady high frame rate to ensure smooth look and feel, while static background is insensitive to instantaneous quality changes. The conventional fixed pipeline cannot distinguish the data, which often results in two extreme cases, namely, high-intensity processing is carried out on all data (including static parts) in order to ensure the quality of dynamic content, so that great waste of computing resources and transmission bandwidth is caused, system power consumption is increased, the processing quality is globally reduced in order to save energy, and text definition or video fluency is inevitably damaged, so that user experience is influenced. Although some improvements exist in the prior art, for example, allowing a user to manually select a "power saving mode" or a "performance mode", this is a coarse-grained global switching, and fine regulation at the frame level and the region level cannot be achieved. There are also some schemes that attempt to switch processing strategies based on simple scene identification (e.g. "video playback application"), but this relies on tags that are not application-layer accurate, and cannot cope with the complications of single-frame intra-mixed content. In addition, other schemes have focused on monitoring the overall load (e.g., CPU utilization) of the system and adjusting display parameters accordingly (e.g., globally decreasing refresh rate), but such adjustments are reactive, global, lack of awareness of the content itself, and may compromise the display quality of all content indifferently while reducing the load. Therefore, the prior art lacks a key mechanism capable of sensing the semantic composition of the display content in real time and finely, and combining the semantic composition with the accurate system running state, so as to implement real self-adaptive and differential processing on different areas at each frame picture level, which is the root cause that the display data processing is difficult to realize dynamic optimal balance among efficiency, quality and power consumption. Disclosure of Invention In view of the above drawbacks of the prior art, an object of the present invention is to provide a method and a system for processing display screen data, which are used for solving the problems of low efficiency and excessive energy consumption caused by the inability of intelligently allocating resources when the display screen processes mixed content. The present invention solves this problem by constructing a closed-loop sensing, decision and execution system. Firstly, analyzing the content characteristics of a picture to be displayed in real time, accurately identifying different areas such as static, dynamic and text, and continuously monitoring the load, bandwidth and power consumption state of the system. Then, the content and the state information are input into a policy decision engine, and the engine performs fusion analysis according to preset optimization rules to dynamically generate a set of compression, rendering and transmission instructions which are customized for different content areas. And finally, the system adaptively schedules bottom hardware resources according to the instructions, and executes differentiated processing flows on various areas. The method enables the system to dynamically adjust the processing strength according to the two key information of what content is and busy system is like an intelligent manager, so that the processing efficiency is maximized and the unnecessary resource consumption is reduced on the premise of guaranteeing the visual experience. The inventi