High-refresh-rate driving method and system of LED display screen

Abstract

The invention relates to the technical field of data processing, and provides a high-refresh-rate driving method and system for an LED display screen, and the method comprises the steps: obtaining to-be-displayed image data, dividing the to-be-displayed image data into a plurality of sub-image frame segments, generating a time sequence change vector data item based on the sub-image frame segments, constructing a two-dimensional map matrix, and carrying out the symbol mapping, thereby obtaining a map symbol sequence, a plurality of map mode index tables are generated to match a corresponding map type for each to-be-refreshed frame in the to-be-displayed image data, and map symbols in the map types are mapped into corresponding row and column scanning data and PWM gray scale control data to refresh the LED display screen in different regions. According to the method, the row and column scanning data and the PWM gray scale control data are generated, and then the control data are used for implementing regional refreshing, so that the refreshing speed of the LED display screen is increased, the transmission delay caused by overlarge data volume in large-screen display is improved, and the problems of low data transmission efficiency and high transmission delay during high-refresh-rate display are solved.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and particularly to a high refresh rate driving method and system for an LED display screen. Background Art

[0002] In recent years, with the continuous progress of display technology and the improvement of users' requirements for picture quality, LED display screens have developed towards high resolution, high brightness, wide color gamut, and high refresh rate. Among them, the refresh rate is regarded as a key indicator determining the stability and smoothness of the display screen. Especially when playing high-speed moving pictures, a high refresh rate can effectively avoid image jitter and trailing phenomena.

[0003] In related technical means, the high refresh rate driving of an LED display screen usually extracts the image data to be displayed into a multi-frame static image sequence; then, for each frame of the image, it performs row-by-row or column-by-column scanning driving, and at the same time, uses pulse width modulation (PWM) control technology to adjust the brightness gray scale values of different pixel points to reproduce a high-definition picture, effectively improving the adaptability of the display screen to complex dynamic pictures and reducing the flicker and instability problems occurring in the full-screen refresh.

[0004] Regarding the above technical solution, although the row-by-row or column-by-column scanning combined with the sub-region refresh mechanism and PWM gray scale control can improve the refresh stability and detail expressiveness of the LED display screen in complex picture scenarios, when displaying at a high refresh rate, the row-by-row and column-by-column scanning driving method will cause a significant increase in data volume with the improvement of the screen resolution, thereby reducing the refresh efficiency; at the same time, it is easily affected by control lag and uneven gray scale distribution, resulting in a blurred or local delay phenomenon in the display picture, and there are problems of low data transmission efficiency and high transmission delay. Summary of the Invention

[0005] In order to improve the problems of low data transmission efficiency and high transmission delay when displaying at a high refresh rate, this application provides a high refresh rate driving method and system for an LED display screen.

[0006] The present invention provides a high refresh rate driving method for an LED display screen, including: acquiring image data to be displayed, dividing the image data to be displayed into a plurality of sub-image frame segments, and generating time series change vector data items based on the sub-image frame segments; constructing a two-dimensional map matrix according to the time series change vector data items, performing symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generating a plurality of map mode index tables based on the map symbol sequence; applying the map mode index tables to match the corresponding map types for each frame to be refreshed in the image data to be displayed, and mapping the map symbols in the map types to corresponding driving gray scale instruction sequences according to a preset symbol gray scale mapping relation table; generating row-column scanning data and PWM gray scale control data according to the driving gray scale instruction sequences, and performing regional refreshing on the LED display screen by applying the row-column scanning data and the PWM gray scale control data.

[0007] As a preferred solution, the steps of acquiring image data to be displayed, dividing the image data to be displayed into a plurality of sub-image frame segments, and generating time series change vector data items based on the sub-image frame segments include: collecting image data to be displayed, dividing the image data to be displayed into a frame sequence set in chronological order, dividing each frame image in the frame sequence set according to a preset pixel grid to obtain a plurality of sub-grids; calculating local gradient change parameters based on the spatial distribution characteristics and gray scale change values of all the sub-grids, and extracting disturbance features according to the edge gray scale structures of all the sub-grids; extracting brightness change trend data and spatial disturbance vectors according to the local gradient change parameters and the disturbance features, and constructing a change structure unit according to the brightness change trend data and the spatial disturbance vectors, where the change structure unit includes a brightness mutation point index and a disturbance offset structure; cross-calibrating the brightness mutation point index and the disturbance offset structure to obtain a set of feature nodes, and dividing sub-image frame segments by using the set of feature nodes; marking change weight information in each sub-image frame segment, and generating time series change vector data items based on the change weight information and the set of feature nodes.

[0008] As a preferred solution, the steps of constructing a two-dimensional map matrix based on the time-series change vector data items, performing symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generating multiple map pattern index tables based on the map symbol sequence include: sorting the time-series change vector data items according to the time series to obtain a time-series change set, performing integration processing on the time-series change set to obtain a brightness change flux sequence, and extracting extreme points of the brightness change flux sequence to generate a trend node set; constructing a brightness trajectory function by using the brightness change flux sequence and the trend node set, generating a periodic jump template structure according to the periodic characteristics of the brightness trajectory function, and extracting peak structure parameters and interval fluctuation parameters based on the brightness trajectory function and the jump template structure; constructing a two-dimensional map matrix according to the peak structure parameters and the interval fluctuation parameters, performing symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, generating structure similarity data according to the distribution pattern of the map symbol sequence in the two-dimensional map matrix; performing joint clustering on the map symbol sequence and the structure similarity data to obtain multiple map pattern clusters, and generating a map pattern index table based on the multiple map pattern clusters, where the map pattern index table records the identifiers, central symbols, and frequency values of each cluster.

[0009] As a preferred solution, the steps of constructing a two-dimensional map matrix according to the peak structure parameters and the interval fluctuation parameters, performing symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generating structure similarity data according to the distribution pattern of the map symbol sequence in the two-dimensional map matrix include: constructing a two-dimensional map matrix by using the peak structure parameters and the interval fluctuation parameters, performing symbolic mapping on each data point in the two-dimensional map matrix with a preset symbol mapping table to obtain a map symbol sequence; extracting the frequency distribution information of each symbol according to the distribution pattern of the map symbol sequence in the two-dimensional map matrix, and calculating the structural similarity between the map symbol sequence and the two-dimensional map matrix based on the frequency distribution information to generate structure similarity data.

[0010] As a preferred solution, the step of matching the corresponding atlas type for each frame to be refreshed in the to-be-displayed image data by using the atlas pattern index table and mapping the atlas symbols in the atlas type to the corresponding driving gray-scale instruction sequence according to the preset symbol gray-scale mapping relation table includes: extracting the corresponding timing change vector data item for each frame to be refreshed from the to-be-displayed image data, using the atlas pattern index table to find the atlas pattern matching the corresponding timing change vector data item, and determining the atlas type to which the frame to be refreshed belongs; selecting the atlas symbol corresponding to the atlas type based on the atlas pattern index table, inputting the atlas symbol corresponding to the atlas type into the preset symbol gray-scale mapping relation table for mapping, and obtaining the corresponding driving gray-scale instruction sequence.

[0011] As a preferred solution, the step of generating row-column scan data and PWM gray-scale control data according to the driving gray-scale instruction sequence and performing sub-region refreshing on the LED display screen by using the row-column scan data and the PWM gray-scale control data includes: reorganizing the driving gray-scale instruction sequence according to pixel regions to generate a set of instruction blocks, constructing a refresh priority matrix based on the change weight data of the instruction block set, performing clustering division on the regions in the refresh priority matrix, and obtaining a regional gray-scale distribution surface; constructing a pulse scheduling queue and a scan scheduling matrix according to the brightness mutation regions in the regional gray-scale distribution surface, synchronously planning the pulse scheduling queue and the scan scheduling matrix, and generating row-column scan data and PWM control data; respectively performing compression encoding on the row-column scan data and the PWM control data to obtain a transmission data group and a dimming mask set, jointly generating a refresh control instruction set by using the transmission data group and the dimming mask set, and performing sub-region refreshing on the LED display screen by using the refresh control instruction set.

[0012] As a preferred solution, the step of dividing the driving gray-scale instruction sequence according to pixel regions to generate a set of instruction blocks, constructing a refresh priority matrix based on the change weight data of the instruction block set, performing clustering division on the regions in the refresh priority matrix, and obtaining a regional gray-scale distribution surface includes: dividing the driving gray-scale instruction sequence according to pixel regions and generating a plurality of instruction blocks based on the division result, summarizing all the instruction blocks to obtain a set of instruction blocks, analyzing the gray-scale change conditions of each instruction block in the instruction block set to obtain gray-scale change information; calculating the corresponding change weight based on the gray-scale change information to obtain the change weight data of each instruction block in the instruction block set; constructing a refresh priority matrix by using the change weight data, performing clustering analysis on the refresh priority matrix, obtaining regions according to the priority values, dividing the regions into different levels according to the priority values to obtain a plurality of refresh priority hierarchical blocks, and generating a regional gray-scale distribution surface based on the refresh priority hierarchical blocks.

[0013] The present application provides a high refresh rate driving system for an LED display screen, including: an acquisition module, configured to acquire image data to be displayed, divide the image data to be displayed into a plurality of sub-image frame segments, and generate timing change vector data items based on the sub-image frame segments; a mapping module, configured to construct a two-dimensional map matrix according to the timing change vector data items, perform symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generate a plurality of map mode index tables based on the map symbol sequence; a matching module, configured to apply the map mode index tables to match the corresponding map type for each frame to be refreshed in the image data to be displayed, and map the map symbols in the map type to the corresponding driving gray scale instruction sequence according to a preset symbol gray scale mapping relationship table; and a control module, configured to generate row-column scanning data and PWM gray scale control data according to the driving gray scale instruction sequence, and perform sub-region refreshing on the LED display screen by applying the row-column scanning data and the PWM gray scale control data.

[0014] Compared with the prior art, the present application has the following beneficial effects: fast transmission efficiency and low transmission delay. By acquiring the image data to be displayed and dividing it into sub-image frame segments, generating timing change vector data items with each sub-image frame segment as a unit, the dynamic change characteristics of the image are described in detail, improving the granularity and fineness of data processing; constructing a two-dimensional map matrix based on the timing change vector data items, performing symbol mapping on the matrix to generate a map symbol sequence, and then generating a plurality of map mode index tables through the map symbol sequence, different picture characteristics can be quickly summarized and classified, facilitating the efficient matching of the map type to which the sub-image frame segment belongs; at the same time, by mapping the map symbols in the matched map type to the corresponding driving gray scale instruction sequence, generating row-column scanning data and PWM gray scale control data based on this, and then using these control data to implement sub-region refreshing, not only significantly improves the refresh speed of the LED display screen, but also effectively solves the transmission delay caused by excessive data volume in large screen display, improving the problems of low data transmission efficiency and high transmission delay in high refresh rate display. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0016] The structures, proportions, sizes, etc. shown in the drawings of this specification are only used to cooperate with the content disclosed in the specification for those skilled in this technology to understand and read, and are not used to limit the conditions for the implementation of the present invention. Therefore, they do not have substantial technical significance. Any modification of the structure, change in the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope that can be covered by the technical content disclosed in the present invention.

[0017] Figure 1 It is a schematic flowchart of the high refresh rate driving method for the LED display screen provided by an embodiment of the present invention; Figure 2 It is a schematic block diagram of the structure of the high refresh rate driving method for the LED display screen provided by an embodiment of the present invention.

[0018] Explanation of reference numerals: 10. High refresh rate driving system for the LED display screen; 11. Acquisition module; 12. Mapping module; 13. Matching module; 14. Control module. Detailed implementation manners

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] The flowchart shown in the drawings is only an example illustration, and does not necessarily include all the content and operations / steps, nor does it necessarily need to be executed in the described order. For example, some operations / steps can also be decomposed, combined, or partially merged. Therefore, the actual execution order may change according to the actual situation.

[0021] It should also be understood that the terms used in this specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this specification of the present application and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0022] It should be further understood that the term "and / or" used in this specification of the present application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0023] Next, the technical solutions of the present invention will be further described in conjunction with the drawings and through specific implementation manners.

[0024] Embodiment 1: As Figure 1 shown, the present application provides a high refresh rate driving method for an LED display screen, including steps S100 to S400.

[0025] Step S100: Obtain the image data to be displayed, divide the image data to be displayed into several sub-image frame segments, and generate time sequence change vector data items based on the sub-image frame segments.

[0026] In this step, when obtaining the image data to be displayed, the signal processing module receives the image data from an external input device, parses the image data, converts a complex dynamic picture into a multi-frame static image sequence, and extracts the pixel information of each static frame. The extracted frames are arranged in chronological order according to the timestamp information, divided into several sub-image frame segments, and time sequence change vector data items are generated for each sub-image frame segment. Specifically, first, each sub-image frame segment after division is split according to a preset pixel grid to form a plurality of sub-grids; then, based on the spatial distribution characteristics and gray scale change values of each sub-grid, local gradient change parameters are calculated, and the gray scale structure characteristics of the sub-grid edges are extracted. These extracted parameters are used to describe the pixel brightness change trend in the sub-image frame segment; subsequently, these features are further analyzed to construct time sequence change vector data items including brightness mutation point indexes, spatial perturbation information, and dynamic change parameters, ensuring that the generated data items can completely represent the time sequence characteristics of the sub-image frame segment.

[0027] For example, for an advertising LED display screen that plays a high-speed dynamic picture, there are 60 frames of video data per second of the picture. According to the frame division principle, each frame of image data will be split into 15 sub-image frame segments, and corresponding time sequence change vector data items are generated by calculating the local gray scale change in each sub-image frame segment, including brightness jump information, spatial offset, and change gradients within different grids. In this way, up to 900 time sequence change vector data items with complete dynamic description features can be generated per second.

[0028] Step S200: Construct a two-dimensional map matrix according to the time sequence change vector data items, perform symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generate a variety of map mode index tables based on the map symbol sequence.

[0029] In this step, the two-dimensional atlas matrix is constructed from time-series change vector data items. Specifically, first, the time-series change vector data items are sorted according to the time dimension to form a complete time-series change set; then, parametric calculations are performed on each time-series change item, including luminance change flux, grayscale difference value, and geometric distribution pattern, so as to form a preliminary two-dimensional data matrix with both spatial and temporal characteristics and smooth it through a filtering algorithm. Next, the symbol mapping rule is applied to each data point in the two-dimensional data matrix, and the data points are converted into atlas symbols with specific semantics according to a preset mapping table, and the symbolization result forms a sequence of atlas symbols, which is used to represent the picture change characteristics of a specific area or time period. Subsequently, by combining the constructed multiple sequences of atlas symbols, cluster analysis is implemented to form multiple different atlas pattern clusters, each cluster corresponding to a pattern category, recording the symbol distribution characteristics and their occurrence frequencies, and finally generating multiple atlas pattern index tables for subsequent atlas type matching.

[0030] For example, when the input video picture contains a dynamic image that scrolls up and down quickly, its luminance change flux fluctuates periodically. When constructing the two-dimensional atlas matrix, the flux value is calculated with the single-frame luminance change of each pixel grid as the input, and the generated atlas symbols include "L" representing luminance mutation points, "S" representing stable regions, "P" representing high-amplitude increase regions, etc. Finally, the generated sequence of atlas symbols is "LSPSSPL", which is drawn as a periodic jump pattern after the index table is generated; while other data contents generate a monotonic change pattern such as "SSSSSSS".

[0031] Step S300: Use the atlas pattern index table to match the corresponding atlas type for each to-be-refreshed frame in the to-be-displayed image data, and map the atlas symbols in the atlas type to the corresponding driving grayscale instruction sequence according to the preset symbol-grayscale mapping relation table.

[0032] In this step, the atlas pattern index table is used to match the atlas type of the input to-be-refreshed frame segment. Specifically, in each to-be-refreshed frame, the corresponding sub-image frame segment is extracted, and the time-series change vector data items generated by the sub-image frame segment are matched with the atlas pattern index table, and the most conforming atlas type identifier is found based on the similarity threshold. Then, through the symbol-grayscale mapping relation table, the atlas symbols included in the matched atlas type are mapped one by one to the driving grayscale instruction sequence. The mapping is completed by looking up the table to ensure the accuracy and high real-time performance of the mapping result. Finally, the generated driving grayscale instruction sequence records the grayscale control information of each pixel point and is used for subsequent row-column scanning and PWM control signal generation.

[0033] For example, a frame of image data contains a periodic jump pattern. After matching with the pattern spectrum index table, it is associated with a pattern with a pattern type ID of "T1", and the corresponding gray-scale control symbol sequence is "100, 150, 200". These symbols are further mapped to a driving gray-scale instruction sequence "IDR: 120, GVR: 180, WBK: 210", and finally row-column scan control data is generated to adjust the brightness value of the LED pixel points.

[0034] Step S400: Generate row-column scan data and PWM gray-scale control data according to the driving gray-scale instruction sequence, and apply the row-column scan data and PWM gray-scale control data to perform sub-region refreshing on the LED display screen.

[0035] In this step, according to the driving gray-scale instruction sequence, the scanning tasks are allocated by region and the row-column scan data for each region is generated. Specifically, according to the preset screen resolution and region setting parameters, the driving gray-scale instruction sequence is split into different sub-blocks, and each sub-block corresponds to a region on the screen. Then, for each sub-block, the row-column scan data is calculated through a scan control algorithm to convert the instruction sequence into a signal stream for operating the LED pixel points. At the same time, in order to accurately control the brightness of different pixel points within the sub-block, a PWM gray-scale control signal is also generated according to the driving gray-scale instruction sequence to determine the duty cycle of the LED diode turn-on time, so as to achieve precise brightness adjustment. Finally, the row-column scan data and PWM gray-scale control signals of each sub-region are sent out synchronously and mapped to the corresponding region of the LED display screen for sub-region refreshing to achieve high-speed and precise display.

[0036] For example, on an LED display screen with a screen resolution of 1920×1080, the entire screen is divided into 4 sub-regions. Each region independently generates row-column scan data "Scan_1 to Scan_4" and PWM data "PWM_1 to PWM_4" according to the instruction sequence. These data are respectively transmitted to the driving modules of each region to implement the sub-region refreshing operation.

[0037] In this embodiment, by obtaining the image data to be displayed, dividing the image data to be displayed into several sub-image frame segments, and generating temporal change vector data items based on the sub-image frame segments to describe the picture brightness change trend and spatial characteristics. Then, a two-dimensional map matrix is constructed according to the temporal change vector data items, and the two-dimensional map matrix is subjected to symbol mapping to obtain a map symbol sequence describing the picture feature distribution. Subsequently, a variety of map pattern index tables are generated based on the map symbol sequence, and the map pattern index tables record typical map types and their characteristics. After that, the map pattern index tables are applied to match the map type to which each frame to be refreshed in the image data to be displayed belongs, and according to a preset symbol gray scale mapping relation table, the map symbols in the map type are further mapped into corresponding driving gray scale instruction sequences. Finally, row-column scan data and PWM gray scale control data are generated according to the driving gray scale instruction sequences, and the row-column scan data and the PWM gray scale control data are applied to perform regional refresh on the LED display screen to complete high-refresh-rate display. By performing fine-grained temporal change analysis on the sub-image frame segments of the image data to be displayed, and introducing a method for constructing and symbol mapping a two-dimensional map matrix based on temporal change vector data items, efficient structured processing of display data and map pattern classification are realized. By generating a variety of map pattern index tables and applying them to match the map type to which the sub-image frame segments belong, the driving gray scale instruction sequences can be quickly generated, thereby effectively improving the generation efficiency of the row-column scan data and the PWM gray scale control data. By performing regional refresh on the LED display screen based on the driving gray scale instruction sequences, not only the real-time dynamic picture performance and detail stability of the high refresh rate are significantly improved, but also the problems of delay and power consumption caused by direct row-by-row or column-by-column scanning are reduced, better meeting the actual needs of high-definition and high-speed display, and improving the problems of low data transmission efficiency and high transmission delay existing in high-refresh-rate display.

[0038] Embodiment 2: In step S100, the image data to be displayed is collected, and the image data to be displayed is divided into a frame sequence set in chronological order, and each frame image in the frame sequence set is sliced according to a preset pixel grid to obtain a number of sub-grids.

[0039] Collecting the image data to be displayed usually involves the display control module obtaining the original picture data from an input signal source (such as a video player, a computer, or a live broadcast signal). The picture data is in RGB format or grayscale dot matrix format. Specifically, the collected picture data will be adjusted to the resolution required by the display screen after input interface adaptation and color space conversion according to the display requirements, and at the same time, it will be sorted into a set of frame sequences in chronological order. After generating the set of frame sequences, each frame of the image is sliced into multiple sub-grids according to a preset pixel grid, and these sub-grids are further refined based on the resolution; for example, if the total resolution of the display screen is 1920×1080 pixels, it can be preset that a single frame of the picture is divided into a 32×18 grid, and each sub-grid contains several pixel points, and the change trends of grayscale and brightness parameters are refined inside the sub-grid. This slicing operation can be achieved through a rectangular partitioning algorithm to ensure that each sub-grid has both spatial continuity and data independence. The sub-grid data after segmentation is sequentially added to a sub-grid queue for subsequent spatial structure analysis and temporal feature processing.

[0040] For example, on an LED display screen with a resolution of 3840×2160, the collected video content is a 1080p dynamic video signal (such as an advertising display or a sports game picture). This video signal is converted into a set of frame sequences through a parsing interface, containing 60 frames per second. Each frame of the picture is then sliced into 60×34 sub-grids through grid slicing, and each sub-grid covers approximately 40 pixel points. Specifically, the data of one frame of the image forms 2040 sub-grids after segmentation, and the data of each sub-grid includes pixel coordinates, brightness values, and grayscale values. These sub-grids are integrated into the sub-grid queue as the input for subsequent grayscale processing.

[0041] Calculate the local gradient change parameters based on the spatial distribution characteristics and grayscale change values of all sub-grids, and extract the perturbation features according to the edge grayscale structure of all sub-grids.

[0042] For all sub-grids, obtain their spatial distribution characteristics through a spatial data analysis algorithm, including the arrangement of adjacent sub-grids, the distribution pattern of pixel points inside the sub-grid, and the global position of the sub-grid within the frame sequence. After obtaining the spatial distribution characteristics of the sub-grids, by scanning the change of grayscale values of sub-grid pixel points one by one, use a gradient calculation formula (such as the Sobel operator) to calculate the local gradient change parameters within the sub-grid to quantify the degree of mutation of brightness and grayscale data. Further, to analyze the edge characteristics of the sub-grid, extract the boundaries of the regions where the grayscale values change drastically in the sub-grid, identify the discontinuous regions of the grayscale structure (i.e., the edge positions), and extract the perturbation features of these edge grayscale structures, including information such as edge length, curvature, and mutation angle, for subsequent construction of the description of the dynamic picture features.

[0043] For example, in the above-mentioned advertisement image with a resolution of 3840×2160, a certain sub-grid contains 50 pixel points, and this sub-grid is located in the upper left corner area of the 10th frame image in the frame sequence. By analyzing its spatial distribution characteristics, the arrangement relationship between this sub-grid and adjacent sub-grids can be determined, and the horizontal and vertical gradient calculation formulas are applied to the gray-scale value changes of its pixel points to obtain local gradient change parameters. Suppose in this sub-grid, the maximum gradient of the brightness change is 50 (unit gray-scale value), while the minimum gradient is 5, then this result indicates that there is a significant brightness jump in this sub-grid. Further, by extracting the gray-scale characteristics of the edge of the sub-grid area, the disturbance characteristics of the gray-scale structure of this sub-grid can be detected, including the boundary curvature of 10°, the length of the jump area accounting for 20% of the total side length, and the brightness mutation direction being diagonally downward to the right.

[0044] Extract the brightness change trend data and spatial disturbance vectors according to the local gradient change parameters and disturbance characteristics, and construct a change structure unit according to the brightness change trend data and spatial disturbance vectors. Among them, the change structure unit includes the brightness mutation point index and the disturbance offset structure.

[0045] Based on the local gradient change parameters and edge gray-scale structure disturbance characteristics of the aforementioned sub-grid, further extract the brightness change trend data inside the sub-grid, use the fitting algorithm to model the time distribution of the brightness gradient value, and obtain the brightness change trend within the sub-grid, including the direction and change rate of brightness enhancement or weakening. At the same time, extract the spatial disturbance vector according to the edge disturbance characteristics to describe the gray-scale dynamic change intensity and position offset attribute within this sub-grid. Combine the extracted brightness change trend data and spatial disturbance vector to construct a change structure unit. Among them, the brightness mutation point index in the change structure unit is used to describe the position of the frame with a rapid brightness jump in the sub-grid, and the disturbance offset structure is used to quantify the direction and amplitude of the disturbance change. This unit is used to further characterize the dynamic characteristics of the image.

[0046] For example, in a sub-grid with a large brightness jump, through the calculation of the gray-scale characteristic data, it is detected that the time curve of the gray-scale change gradient shows an upward trend, and the brightness rapidly increases from 50 to 200 within the 8th to 12th frames. The extracted brightness change trend data is: [Growth direction: positive, change rate: 30 gray-scale values / frame]. At the same time, the offset vector of the edge disturbance characteristics describes the tilt direction (upper right offset) and intensity (jump amplitude) of the gray scale during the jump; the change structure unit synthesized by these two records (Brightness mutation point index position: the 8th frame; Offset structure: Offset direction upper right, amplitude 5 pixels). These structure data provide a dynamic reference for the subsequent sub-image frame segment analysis and refreshing.

[0047] Cross-calibrate the brightness mutation point index and the disturbance offset structure to obtain a set of characteristic nodes, and use the set of characteristic nodes to divide the sub-image frame segments.

[0048] Combine the indexes of the luminance mutation points extracted within the sub-grid with the spatial perturbation offset structure. Through cross-analysis methods (such as coordinate superposition algorithms), map the positions of the mutation points onto the time axis of the frame sequence to form a set of characteristic nodes for the sub-grid. Meanwhile, based on the position distribution of the set of characteristic nodes, construct a time-space distribution model for the frame segments, and use this to divide the sub-image frame segments. Specifically, the division of the frame segments needs to comprehensively consider the correlation between the characteristic nodes among the frames, ensuring that the temporal changes described by each frame segment are independent and contain the most significant parts of the dynamic changes of the sub-grid, thereby providing a time slice with complete features for further analysis.

[0049] For example, in a certain advertisement picture, the gray-scale mutations in the sub-grid data between the 12th frame and the 20th frame are most frequent in the first quadrant area. By cross-analyzing the luminance mutation points and the perturbation offset structure, two sets of key characteristic nodes can be obtained: {12th frame, (100, 200), offset 5 pixels to the upper right} and {18th frame, (150, 250), offset 8 pixels to the lower right}. Divide the 12th frame to the 20th frame into a sub-image frame segment to describe the dynamic features of this segment of the picture.

[0050] Annotate the change weight information in each sub-image frame segment, and generate time-series change vector data items based on the change weight information and the set of characteristic nodes.

[0051] In each sub-image frame segment, weight-label the degree of dynamic change of each characteristic region within the frame segment. The weight value is positively correlated with the gray-scale change rate, the number of luminance mutations, and the perturbation amplitude. Subsequently, based on the weight-labeling information and the set of characteristic nodes, comprehensively extract the key time-series features within a single frame segment, and construct time-series change vector data items. These data items contain multi-dimensional dynamic features, including spatial distribution changes, luminance change distribution on the time axis and its continuity description, and finally provide dynamic inputs for the subsequent construction of the two-dimensional map matrix.

[0052] For example, in a sub-image frame segment, it contains 10 frames of pictures, and its change weight (calculated based on the number and intensity of mutations) values are: {1st frame: 0.8, 2nd frame: 0.9, 8th frame: 1.5, 10th frame: 1.0}; the resulting time-series change vector data item is [{frame index: 1 - 10, luminance change: linearly increasing (rate 50 / frame), perturbation direction: offset to the upper right, weight annotation: 1.0 (high)}].

[0053] In step S200, sort the time-series change vector data items according to the time series to obtain a time-series change set, perform integral processing on the time-series change set to obtain a luminance change flux sequence, and extract the extreme points of the luminance change flux sequence to generate a set of trend nodes.

[0054] Sort the time - series change vector data items according to the frame numbers of the time series to form a time - series change set, ensuring that the time - series change data of each frame is arranged in order in the set. Specifically, calculate the brightness change amount for each time - series change vector data item, and accumulate through an integration algorithm (such as numerical integration method) to obtain a brightness change flux sequence, where the brightness change flux sequence can accurately reflect the overall change trend of the picture brightness factor over time. In the flux sequence, further extract the extreme points of the brightness change through an extreme - value detection algorithm, including local maxima and local minima, and use these extreme points as trend nodes to describe the significant change positions and time points of the picture brightness. These trend nodes can serve as an important basis for classifying picture features and provide a foundation for constructing the brightness trajectory function subsequently.

[0055] For example, assume that a dynamic advertisement picture contains 10 consecutive frames, and the brightness change values are: {50, 70, 90, 110, 130, 150, 130, 110, 70, 40}. The time - series change set formed by time sorting is the frame sequence {frame 1 to frame 10}, and the corresponding brightness change flux sequence is obtained through integral calculation: {50, 120, 210, 320, 450, 600, 730, 840, 910, 950}. In the flux sequence, the extreme points are extracted as follows: local maximum (brightness 600 at frame 6), local minima (brightness 50 at frame 1 and brightness 950 at frame 10), and the generated set of trend nodes is {{frame 1, 50}, {frame 6, 600}, {frame 10, 950}}.

[0056] Construct a brightness trajectory function using the brightness change flux sequence and the set of trend nodes, generate a periodic jump template structure based on the periodic characteristics of the brightness trajectory function, and extract peak structure parameters and interval fluctuation parameters based on the brightness trajectory function and the jump template structure.

[0057] According to the brightness change flux sequence and the set of trend nodes, use a curve - fitting algorithm (such as cubic spline interpolation method) to construct a brightness trajectory function, which is used to describe the brightness fluctuation trend of the picture over time. In particular, the brightness trajectory function contains the change and fluctuation pattern and continuity characteristics of the picture brightness. After constructing the brightness trajectory function, extract its periodic characteristics through frequency - domain analysis, and generate a periodic jump template structure based on the detected period information. This template structure can capture the main period and amplitude change characteristics of the picture jump. Further analyze the positions of the peaks and valleys of the brightness trajectory function to extract peak structure parameters and interval fluctuation parameters, where the peak structure parameters include peak height and valley depth, and the interval fluctuation parameters describe the time interval between peaks and valleys. The structure parameters output in this part lay a data foundation for establishing a two - dimensional map matrix subsequently.

[0058] For example, for the luminance trajectory function constructed from the 10-frame data in the above example, the fitting result is f(t) = 50 + 100sin(πt / 5). According to this trajectory function, it is detected that it has a periodic feature with a period of 5 frames, and its periodic jump template structure can be defined as {period = 5 frames, fluctuation amplitude = 100}. Further, its peak value (luminance of 150 in the 6th frame), trough value (luminance of 50 in the 1st frame and luminance of 40 in the 10th frame), and the time interval between the peak and trough (5 frames) are extracted.

[0059] Construct a two-dimensional map matrix based on the peak structure parameters and interval fluctuation parameters, perform symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generate structure similarity data according to the distribution pattern of the map symbol sequence in the two-dimensional map matrix.

[0060] Using the peak structure parameters and interval fluctuation parameters extracted above, construct a preliminary two-dimensional map matrix according to the spatial and temporal dimensions. Specifically, taking the time interval between the peak and trough as the horizontal axis and the fluctuation amplitude or luminance difference as the vertical axis, fill the luminance change characteristics of each time node into the corresponding matrix cell to form a two-dimensional data array indicating the ups and downs of the picture luminance. Subsequently, according to a preset symbol mapping table, perform symbolic mapping processing on each data point in the two-dimensional map matrix, map the luminance difference value or change frequency value to a discrete symbol (such as "L" representing low luminance, "H" representing high luminance, "S" representing a stable area, etc.), and generate a map symbol sequence through mapping. Further, according to the distribution pattern of the generated map symbol sequence in the two-dimensional map matrix, calculate the frequency distribution information of each symbol, and combine the frequency distribution data to analyze the spatial relationship and structural similarity between the symbols, and generate accurate structure similarity data to describe the global characteristics of the map symbol layout.

[0061] For example, for the periodic jump template structure {period = 5 frames, fluctuation amplitude = 100} and the peak and trough data of the luminance trajectory function, construct a two-dimensional map matrix as shown in Table 1.

[0062] Table 1: The mapped map symbol sequence is {LHHLLHHLL}, the frequency distribution between the symbols is {L: 4, H: 4}, and symmetric structure similarity data is generated in combination with the frequency data to describe the periodic change rule contained in this matrix.

[0063] Among them, the steps of constructing a two-dimensional map matrix according to the peak structure parameters and the interval fluctuation parameters, performing symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generating structure similarity data according to the distribution pattern of the map symbol sequence in the two-dimensional map matrix include: constructing a two-dimensional map matrix by using the peak structure parameters and the interval fluctuation parameters, and performing symbolic mapping on each data point in the two-dimensional map matrix with a preset symbol mapping table to obtain a map symbol sequence.

[0064] After constructing a two-dimensional map matrix according to the peak structure parameters, each data point in the matrix cell is input into a preset symbol mapping table for mapping. This mapping table discretizes the matrix cell values into several symbol types through predefined rules. For example, the brightness difference of 0-50 can be marked as "L", the brightness difference of 50-100 can be marked as "M", and the brightness difference above 100 can be marked as "H". After symbolization, the original matrix will be transformed into a data stream containing a significant feature symbol sequence for further analysis of its distribution law.

[0065] For example, for the original matrix values {30, 80, 110}, the symbol sequence obtained through the mapping table conversion is "LMH".

[0066] According to the distribution pattern of the map symbol sequence in the two-dimensional map matrix, extract the frequency distribution information of each symbol, and calculate the structure similarity between the map symbol sequence and the two-dimensional map matrix based on the frequency distribution information to generate structure similarity data.

[0067] By analyzing the spatial distribution pattern of the map symbol sequence in the two-dimensional map matrix, calculate the frequency distribution value of each symbol in the matrix. For example, for the "L" and "H" symbol parts of the same sequence, respectively count their occurrence times and spatial concentration, and calculate the similarity score between the symbol sequence and the whole matrix through cosine similarity or Mahalanobis distance to generate structure similarity data describing power-law distribution or periodic characteristics.

[0068] For example, if the map symbol sequence is {L, L, H, H}, the occurrence frequency is 2 / 4, and the distribution concentration is 25%, the finally generated structure similarity is displayed as: {low difference degree matching}.

[0069] Perform joint clustering on the map symbol sequence and the structure similarity data to obtain multiple map pattern clusters, and generate a map pattern index table based on the multiple map pattern clusters. Among them, the map pattern index table records the identification, central symbol, and frequency value of each cluster.

[0070] Taking the generated atlas symbol sequence and the corresponding structural similarity data as input, use a joint clustering algorithm (such as K-means, hierarchical clustering, DBSCAN, etc.) to classify the data. Specifically, represent the symbol sequence as a high-dimensional feature vector, and combine the frequency distribution information and spatial relationship in the structural similarity data to define a similarity metric standard for clustering (such as cosine similarity or Euclidean distance). After clustering, each class of symbol sequences represents an atlas pattern cluster, and the central symbol of each cluster is the symbol sequence with the strongest representativeness in the cluster. For example, by selecting the symbol with the highest frequency or the highest similarity as the cluster center, the central symbol can be determined. Further statistically analyze the symbol occurrence frequency of each cluster in the clustering result, and finally generate an atlas pattern index table. The atlas pattern index table records the identifier of each cluster, the central symbol, and its frequency value, which is used to quickly retrieve the most matching atlas pattern cluster.

[0071] For example, assume that three groups of atlas symbol sequences are generated in the display screen: {LHHLHHLL} (frequency 10 times), {HHHHLLLL} (frequency 7 times), {LHLLLHLH} (frequency 5 times). After joint clustering by Euclidean distance, three atlas pattern clusters are obtained: cluster 1 (with {LHHLHHLL} as the central symbol and a frequency value of 10), cluster 2 (with {HHHHLLLL} as the central symbol and a frequency value of 7), and cluster 3 (with {LHLLLHLH} as the central symbol and a frequency value of 5).

[0072] In step S300, extract the time series change vector data item corresponding to each frame to be refreshed from the image data to be displayed, and use the atlas pattern index table to find the atlas pattern that matches the corresponding time series change vector data item, and determine the atlas type to which the frame to be refreshed belongs.

[0073] Extract the time series change vector data item corresponding to each frame to be refreshed from the image data to be displayed to describe the dynamic change characteristics of the frame, and match this data item with the previously generated atlas pattern index table. Specifically, extract the brightness change trend, symbol sequence distribution characteristics, and their corresponding structural similarity information in the time series change vector data item. Combining the central symbol and frequency value recorded in the atlas pattern index table, the atlas pattern cluster most similar to this data item can be found through cosine similarity or other matching algorithms, and then the atlas type to which the frame to be refreshed belongs can be determined. If multiple atlas types with small differences appear simultaneously, the type with a high frequency value is preferentially matched to ensure stability and accuracy.

[0074] For example, assume that the symbol sequence generated by the timing change vector data item corresponding to the frame to be refreshed is {LHHLHHLL}, its brightness change range is (50 - 150), and the structural similarity score is 0.92. In the atlas pattern index table, the cluster most similar to this symbol sequence is cluster 1 (central symbol: LHHLHHLL, frequency value is 10), and the matching result indicates that the atlas type of the frame to be refreshed is "atlas type 1 (cluster 1)". Therefore, the frame to be refreshed is classified as atlas type 1.

[0075] Select the atlas symbols corresponding to the atlas type based on the atlas pattern index table, and input the atlas symbols corresponding to the atlas type into the preset symbol - gray - scale mapping relation table for mapping to obtain the corresponding driving gray - scale instruction sequence.

[0076] Based on the atlas type determined in the previous step, extract the atlas symbols corresponding to this type from the atlas pattern index table. Subsequently, input the extracted atlas symbols into the preset symbol - gray - scale mapping relation table for mapping. The symbol - gray - scale mapping relation table maps each atlas symbol to a specific gray - scale control instruction through mapping rules. The gray - scale control instruction contains the brightness setting value of the pixel point and the time distribution information, thus forming a driving gray - scale instruction sequence. This instruction sequence records the brightness adjustment parameters and the underlying parameters of the PWM control signal for each pixel area in the frame to be refreshed, providing a basis for subsequent row - column scanning control.

[0077] For example, for the aforementioned "atlas type 1" to which the frame to be refreshed belongs, its central symbol is {LHHLHHLL}. Through the symbol - gray - scale mapping relation table, map the symbol "L" to the gray - scale value 50 and the symbol "H" to the gray - scale value 150. Therefore, the generated driving gray - scale instruction sequence is {50, 150, 150, 50, 150, 150, 50, 50}. This sequence provides accurate and direct gray - scale control signals for each small block of pixel points.

[0078] In step S400, reorganize the driving gray - scale instruction sequence by pixel area to generate a set of instruction blocks, and construct a refresh priority matrix based on the change weight data of the instruction block set. Cluster and divide the regions in the refresh priority matrix to obtain the regional gray - scale distribution surface.

[0079] Convert the driving grayscale instruction sequence from the overall pixel distribution to data blocks divided by pixel regions, and perform partition recombination operations. Specifically, the resolution of the image to be displayed is divided into several fixed pixel regions (for example, with 16×16 pixels as a unit), and an instruction block is generated for each region, which contains the grayscale control information of all pixels in the block. Summarize all the instruction blocks to form an instruction block set. For each summarized instruction block, by analyzing its brightness change characteristics (such as the dynamic change range of grayscale, average grayscale value, etc.), calculate the grayscale change information of each instruction block. The grayscale change information is used to reflect the dynamic change degree of the instruction block area and provide a basis for subsequent scheduling and refresh optimization. Further statistically analyze the grayscale change information of the instruction block set to generate change weight data. The change weight data is used to allocate the priority of each instruction block when concentrating refresh resources. Regions with higher priority will obtain more resource allocation during the refresh process. Use this change weight data to construct a refresh priority matrix for sub-region refresh optimization of the display screen.

[0080] For example, on a 1920×1080 LED display screen, the entire screen is divided into 120×30 regions of 16×16 pixels, and each region contains 256 pixels. For a dynamically changing advertising screen, 3600 instruction block set data is generated through the instruction block summarization operation. Suppose the grayscale range of a certain instruction block area is from 50 to 200 from frame 1 to frame 10, then its grayscale change information is marked as "strong dynamic", and the corresponding change weight value is set to a higher priority; if the grayscale difference in another region is only between 50 and 55, it is marked as "weak dynamic", and the refresh resource occupancy priority is lower.

[0081] Among them, the steps of reorganizing the driving grayscale instruction sequence by pixel region, generating an instruction block set, constructing a refresh priority matrix based on the change weight data of the instruction block set, and clustering and dividing the regions in the refresh priority matrix to obtain the regional grayscale distribution surface include: dividing the driving grayscale instruction sequence by pixel region, generating multiple instruction blocks based on the division result, summarizing all the instruction blocks to obtain an instruction block set, and analyzing the grayscale change situation of each instruction block in the instruction block set to obtain the grayscale change information.

[0082] Divide the driving grayscale instruction sequence into several fixed regions according to the resolution of the display screen, extract the pixel grayscale instructions of each region, and generate corresponding multiple instruction blocks. The division method of the instruction blocks is usually based on a rectangular grid. For example, a screen with a resolution of 1920×1080 can be divided into small fixed grids (such as 16×16 pixels). Each instruction block data records the pixel brightness, grayscale modulation value, and dynamic change parameters within the grid. Subsequently, summarize all the divided instruction blocks to form a global instruction block set. To describe the dynamic change behavior of the instruction block region, perform a grayscale change analysis operation on each instruction block. Specifically, by calculating indicators such as the grayscale range, grayscale change rate, and grayscale jump frequency of the instruction block, extract the grayscale change information of this region to provide a reference for subsequent processing.

[0083] For example, assume that a display screen with a resolution of 3840×2160 is divided into 256×135 instruction blocks of 16×16, and the grayscale of a certain region (assumed to be the grid at the 20th row and 15th column) changes as follows from frame 1 to frame 10: the average grayscale of frame 1 is 50, the average grayscale of frame 10 is 200, and the grayscale jump frequency is 5 times / frame. Then the grayscale change information is {grayscale range: 150, change rate: 15 grayscales / frame, jump frequency: 5 times / frame}.

[0084] Calculate the corresponding change weights based on the grayscale change information to obtain the change weight data of each instruction block in the instruction block set.

[0085] Calculate the corresponding change weight for each instruction block using the grayscale change information of each instruction block. Specifically, comprehensively consider indicators such as the grayscale range, dynamic jump frequency, and change time period of the instruction block, and generate the change weight value of each instruction block according to the weighting rule (such as linear weight summation). The higher the change weight, the more significant the dynamic change of the region corresponding to the instruction block on the display screen, so it needs to be refreshed and adjusted preferentially. Further normalize the change weight values, map all weight values to the [0,1] interval to form a unified change weight data set, and append it to the corresponding instruction block set.

[0086] For example, assume that the data of three instruction blocks are respectively: Grayscale range = 100, jump frequency = 10 -> change weight = 0.8; Grayscale range = 50, jump frequency = 5 -> change weight = 0.5; Grayscale range = 10, jump frequency = 2 -> change weight = 0.2.

[0087] After normalization, these weight values still maintain the proportional relationship, which is {0.8, 0.5, 0.2}, and the distribution weight data of all instruction blocks are added to the instruction block set.

[0088] Construct a refresh priority matrix using variable weight data, perform clustering analysis on the refresh priority matrix, obtain regions according to the priority values, divide the regions into different levels according to the priority values to get multiple refresh priority stratified blocks, and generate a regional gray-scale distribution surface based on the refresh priority stratified blocks.

[0089] Use the variable weight data as input to construct a refresh priority matrix. Each matrix cell corresponds to a pixel region on the screen, and its value is the variable weight of that region. After the matrix is constructed, use a clustering analysis algorithm (such as K-means, etc.) to classify the priority values in the matrix, and divide the regions with similar variable weights into the same priority level. Through this classification, multiple "refresh priority stratified blocks" are output, and the update frequency or resource allocation degree of each block is determined by the priority value. On this basis, further statistically analyze the gray-scale change distribution within each hierarchical block to generate a more refined regional gray-scale distribution surface for controlling the gray-scale change parameters of each region.

[0090] For example, after clustering analysis, it can be divided into 2 stratified blocks, where the region with a weight value of 0.6 is marked as a high-priority block, and the region with a weight value of 0.3 is marked as a low-priority block.

[0091] Construct a pulse scheduling queue and a scan scheduling matrix based on the brightness mutation regions in the regional gray-scale distribution surface, perform synchronous planning on the pulse scheduling queue and the scan scheduling matrix to generate row-column scan data and PWM control data.

[0092] In the regional gray-scale distribution surface, identify the mutation regions with significant brightness jumps, and construct a pulse scheduling queue according to the priorities of the mutation regions to control the refresh order of these regions. At the same time, establish a scan scheduling matrix to define the scheduling rules for the row-column signals of the display screen. By performing synchronous planning on the pulse scheduling queue and the scan scheduling matrix, ensure that the refresh operations of all regions are coordinated and the refresh efficiency is maximally improved. The adjusted scheduling signals are consistent with the display requirements, providing row-column scan data and PWM gray-scale control data for the LED display screen.

[0093] For example, if the high-priority regions in the regional gray-scale distribution surface are distributed in the grid formed by rows 1 - 16 and columns 1 - 16, the pulse scheduling queue will take this region as the first-level scheduling block and allocate higher refresh resources.

[0094] Perform compression encoding on the row-column scan data and the PWM control data respectively to obtain a transmission data group and a dimming mask set, jointly generate a refresh control instruction set from the transmission data group and the dimming mask set, and use the refresh control instruction set to perform regional refresh on the LED display screen.

[0095] The generated row-column scan data and PWM grayscale control data are compressed, and the transmission overhead is reduced through a lossless compression algorithm. The generated compressed encoded data is uniformly stored in the transmission data group. Meanwhile, a dimming mask set is constructed according to the PWM data for precisely controlling the details of the brightness modulation of each pixel. The transmission data group and the dimming mask set are jointly used to generate a refresh control instruction set, which records the complete data stream for full-screen update. By calling the control module, the refresh control instruction set is used to perform sub-region refresh operations on regions with different priorities, achieving high-refresh-rate display.

[0096] For example, an instruction set contains 300 bytes of compressed scan data and 150 bytes of PWM control data, and the total distribution is scheduled to the control and drive module of the LED display according to different priorities, realizing the smooth refresh of dynamic images.

[0097] In this embodiment, by performing refined analysis on the time-sequence change vector data items of the image data to be displayed, combining the grayscale change, spatial distribution characteristics and their dynamic characteristics, a method of constructing a two-dimensional spectrum matrix and symbol mapping is adopted to successfully convert the dynamic image data into an efficient spectrum pattern index table. By matching the time-sequence change vector with the spectrum pattern, the spectrum type of the frame to be refreshed is accurately determined, ensuring the reasonable allocation of the spectrum type during image refresh. Further, in this embodiment, through clustering analysis and change weight calculation, the priority scheduling during the image refresh process is optimized, a fine-grained refresh priority matrix is constructed, and a regional grayscale distribution surface is generated, improving the refresh efficiency and accuracy. On this basis, combined with the synchronous planning of the pulse scheduling queue and the scan scheduling matrix, the row-column scan data and PWM control data are generated. Finally, the transmission data is optimized through compression encoding, reducing the transmission cost of the refresh control instruction set, and realizing the sub-region refresh of the LED display. The overall solution significantly improves the smoothness and stability of the display effect through efficient data processing, dynamic adjustment and precise control. Especially under the high-refresh-rate requirement, it optimizes the image detail performance and resource scheduling, meeting the requirements of complex dynamic image display.

[0098] Embodiment 3: As Figure 2 shown, the present application also provides a high-refresh-rate driving system 10 for an LED display, including an acquisition module 11, a mapping module 12, a matching module 13 and a control module 14.

[0099] The acquisition module 11 is mainly used to acquire the image data to be displayed, divide the image data to be displayed into several sub-image frame segments, and generate time-sequence change vector data items based on the sub-image frame segments.

[0100] The mapping module 12 is mainly used to construct a two-dimensional atlas matrix based on the time-series change vector data items, perform symbol mapping on the two-dimensional atlas matrix to obtain an atlas symbol sequence, and generate multiple atlas pattern index tables based on the atlas symbol sequence.

[0101] The matching module 13 is mainly used to apply the atlas pattern index table to match the corresponding atlas type for each to-be-refreshed frame in the to-be-displayed image data, and map the atlas symbols in the atlas type to the corresponding drive gray-scale instruction sequence according to the preset symbol gray-scale mapping relation table.

[0102] The control module 14 is mainly used to generate row-column scan data and PWM gray-scale control data according to the drive gray-scale instruction sequence, and apply the row-column scan data and PWM gray-scale control data to perform regional refresh on the LED display screen.

[0103] In this embodiment, through the collaborative work of the acquisition module 11, the mapping module 12, the matching module 13, and the control module 14, a complete high-refresh-rate drive system is formed from the acquisition of image data to the regional refresh of the display screen. The acquisition module 11 generates time-series change vector data items that can represent the characteristics of dynamic pictures through frame serialization processing and spatial segmentation of the to-be-displayed image data, providing the basic input for subsequent data processing. The mapping module 12 constructs a two-dimensional atlas matrix based on the time-series change vector data items, generates an atlas symbol sequence with global feature distribution through symbol mapping, and further generates multiple atlas pattern index tables, which can efficiently classify and store the feature patterns of different dynamic pictures. The matching module 13 uses the generated atlas pattern index table to accurately match the corresponding atlas type for each to-be-refreshed picture, and through the symbol gray-scale mapping relation table, converts the symbols of the matched atlas type into a drive gray-scale instruction sequence, providing precise gray-scale adjustment data for drive control. The control module 14 generates row-column scan data and PWM gray-scale control data according to the drive gray-scale instruction sequence, and dynamically adjusts the refresh resource allocation of each area of the display screen through regional refresh operations, optimizing the refresh performance of the brightness mutation area, and realizing the stability and smoothness of complex pictures and high-refresh-rate displays. The high-refresh-rate drive system provided in this embodiment realizes the organic combination of data processing, dynamic matching, and precise control through modular design and functional division, greatly improving the display effect of the display screen in high-dynamic pictures, while reducing the transmission delay and power consumption, and meeting the actual needs of high-definition and high-speed display scenarios.

[0104] It should be noted that those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described system and each module can refer to the corresponding processes in the foregoing Embodiment 1, and will not be elaborated herein.

[0105] The structures, proportions, sizes, etc. shown in the accompanying drawings of this specification are only used to cooperate with the content disclosed in the specification for those familiar with this technology to understand and read, and are not used to limit the limiting conditions under which the present invention can be implemented. Therefore, they do not have any substantial technical significance. Any modification of the structure, change in the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope that can be covered by the technical content disclosed in the present invention.

[0106] As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high refresh rate driving method for an LED display screen, characterized in that, Including: Obtain the image data to be displayed, divide the image data to be displayed into several sub-image frame segments, and generate time-series change vector data items based on the sub-image frame segments; Construct a two-dimensional map matrix according to the time-series change vector data items, perform symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generate multiple map pattern index tables based on the map symbol sequence; Apply the map pattern index table to match the corresponding map type for each frame to be refreshed in the image data to be displayed, and map the map symbols in the map type to the corresponding drive gray-scale instruction sequence according to a preset symbol gray-scale mapping relationship table; Generate row-column scan data and PWM gray-scale control data according to the drive gray-scale instruction sequence, and perform sub-region refreshing on the LED display screen by applying the row-column scan data and the PWM gray-scale control data.

2. The high refresh rate driving method of the LED display screen according to claim 1, wherein, The steps of obtaining the image data to be displayed, dividing the image data to be displayed into several sub-image frame segments, and generating time-series change vector data items based on the sub-image frame segments include: Collect the image data to be displayed, divide the image data to be displayed into a frame sequence set in chronological order, and slice each frame image in the frame sequence set according to a preset pixel grid to obtain several sub-grids; Calculate local gradient change parameters based on the spatial distribution characteristics and gray-scale change values of all the sub-grids, and extract perturbation features according to the edge gray-scale structures of all the sub-grids; Extract brightness change trend data and spatial perturbation vectors according to the local gradient change parameters and the perturbation features, and construct a change structure unit according to the brightness change trend data and the spatial perturbation vectors, where the change structure unit includes a brightness mutation point index and a perturbation offset structure; Cross-calibrate the brightness mutation point index and the perturbation offset structure to obtain a set of feature nodes, and use the set of feature nodes to divide sub-image frame segments; Mark change weight information in each sub-image frame segment, and generate time-series change vector data items based on the change weight information and the set of feature nodes.

3. The high refresh rate driving method of the LED display screen according to claim 1, characterized in that The steps of constructing a two-dimensional map matrix according to the time-series change vector data items, performing symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generating multiple map pattern index tables based on the map symbol sequence include: Sort the time-series change vector data items according to the time series to obtain a time-series change set, perform integration processing on the time-series change set to obtain a brightness change flux sequence, and extract extreme points of the brightness change flux sequence to generate a set of trend nodes; Construct a brightness trajectory function by using the brightness change flux sequence and the set of trend nodes, generate a periodic jump template structure according to the periodic characteristics of the brightness trajectory function, and extract peak structure parameters and interval fluctuation parameters based on the brightness trajectory function and the jump template structure; Construct a two-dimensional map matrix according to the peak structure parameters and the interval fluctuation parameters, perform symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generate structure similarity data according to the distribution pattern of the map symbol sequence in the two-dimensional map matrix; Perform joint clustering on the map symbol sequence and the structure similarity data to obtain multiple map pattern clusters, and generate a map pattern index table based on the multiple map pattern clusters, where the map pattern index table records the identifiers, central symbols, and frequency values of each cluster.

4. The high refresh rate driving method of the LED display screen according to claim 3, characterized in that, The steps of constructing a two-dimensional map matrix according to the peak structure parameters and the interval fluctuation parameters, performing symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generating structure similarity data according to the distribution pattern of the map symbol sequence in the two-dimensional map matrix include: Construct a two-dimensional map matrix using the peak structure parameters and the interval fluctuation parameters, and perform symbolic mapping on each data point in the two-dimensional map matrix with a preset symbol mapping table to obtain a map symbol sequence; According to the distribution pattern of the map symbol sequence in the two-dimensional map matrix, extract the frequency distribution information of each symbol, and calculate the structure similarity between the map symbol sequence and the two-dimensional map matrix based on the frequency distribution information to generate structure similarity data.

5. The high refresh rate driving method of the LED display screen according to claim 1, characterized in that The steps of applying the map pattern index table to match the corresponding map type for each to-be-refreshed frame in the to-be-displayed image data, and mapping the map symbols in the map type to the corresponding driving grayscale instruction sequence according to a preset symbol grayscale mapping relationship table include: Extract the corresponding timing change vector data item for each to-be-refreshed frame from the to-be-displayed image data, and use the map pattern index table to find the map pattern that matches the corresponding timing change vector data item to determine the map type to which the to-be-refreshed frame belongs; Select the map symbol corresponding to the map type based on the map pattern index table, and input the map symbol corresponding to the map type into a preset symbol grayscale mapping relationship table for mapping to obtain the corresponding driving grayscale instruction sequence.

6. The high refresh rate driving method of the LED display screen according to claim 1, wherein, The steps of generating row-column scan data and PWM grayscale control data according to the driving grayscale instruction sequence, and performing regional refresh on the LED display screen using the row-column scan data and the PWM grayscale control data include: Reorganize the driving grayscale instruction sequence by pixel region to generate a set of instruction blocks, and construct a refresh priority matrix based on the change weight data of the set of instruction blocks. Perform clustering division on the regions in the refresh priority matrix to obtain a regional grayscale distribution surface; Construct a pulse scheduling queue and a scan scheduling matrix according to the brightness mutation regions in the regional grayscale distribution surface, and perform synchronous planning on the pulse scheduling queue and the scan scheduling matrix to generate row-column scan data and PWM control data; Perform compression encoding on the row-column scan data and the PWM control data respectively to obtain a transmission data group and a dimming mask set, jointly generate a refresh control instruction set from the transmission data group and the dimming mask set, and use the refresh control instruction set to perform area-by-area refresh on the LED display screen.

7. The high refresh rate driving method of the LED display screen according to claim 6, wherein The step of reorganizing the driving gray-scale instruction sequence by pixel area to generate an instruction block set, constructing a refresh priority matrix based on the change weight data of the instruction block set, and performing clustering division on the areas in the refresh priority matrix to obtain an area gray-scale distribution surface includes: Divide the driving gray-scale instruction sequence by pixel area, generate a plurality of instruction blocks based on the division result, summarize all the instruction blocks to obtain an instruction block set, and analyze the gray-scale change situation of each instruction block in the instruction block set to obtain gray-scale change information; Calculate the corresponding change weight based on the gray-scale change information to obtain the change weight data of each instruction block in the instruction block set; Apply the change weight data to construct a refresh priority matrix, perform clustering analysis on the refresh priority matrix, obtain regions according to the priority value, divide the regions into different levels according to the priority value to obtain a plurality of refresh priority hierarchical blocks, and generate an area gray-scale distribution surface based on the refresh priority hierarchical blocks.

8. A high refresh rate driving system for an LED display screen, characterized in that, Include: An acquisition module, configured to acquire the image data to be displayed, divide the image data to be displayed into several sub-image frame segments, and generate a timing change vector data item based on the sub-image frame segments; A mapping module, configured to construct a two-dimensional map matrix according to the timing change vector data item, perform symbol mapping on the two-dimensional map matrix to obtain a map symbol sequence, and generate a plurality of map pattern index tables based on the map symbol sequence; A matching module, configured to apply the map pattern index table to match the corresponding map type for each frame to be refreshed in the image data to be displayed, and map the map symbols in the map type to the corresponding driving gray-scale instruction sequence according to a preset symbol gray-scale mapping relationship table; A control module, configured to generate row-column scan data and PWM gray-scale control data according to the driving gray-scale instruction sequence, and perform area-by-area refresh on the LED display screen by using the row-column scan data and the PWM gray-scale control data.

Images