Image data enhancement method and system applied to maneuvering command screen

Abstract

The invention provides an image data enhancement method and system applied to a maneuvering command screen, and relates to the technical field of image data processing, and the method comprises the steps: firstly obtaining to-be-enhanced image data, command scene real-time semantic data and enhancement effect feedback data of the maneuvering command screen; constructing command scene semantic association features; generating an enhancement strategy self-adaptive iteration rule based on the command scene semantic association features; calling a semantic perception enhancement model to execute collaborative operation to generate preliminarily enhanced image data; feedback information is collected to update semantic association features and iteration rules, and secondary enhancement is carried out on the preliminary enhanced image data to generate final enhanced image data; and finally, transmitting the enhanced image data to a display driving unit, and storing related information to form an associated record set, so that the image data quality and applicability of the maneuvering command screen can be effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image data processing, in particular to an image data enhancement method and system applied to a mobile command screen. BACKGROUND

[0002] In a mobile command scene, as the core device for information display and interaction, the display quality of the image data of the mobile command screen directly affects the accuracy and efficiency of command decision-making. With the increasing complexity and dynamics of command tasks, the requirements for the image data displayed by the mobile command screen are also increasing.

[0003] Currently, most of the methods for image data enhancement focus on the general image field, mainly focusing on the optimization of basic visual features such as image clarity and color saturation. However, the image data displayed by the mobile command screen has its own characteristics, not only containing command situation interactive image elements and instruction operation text image elements, but also being affected by real-time changes in the command scene, such as the advancement of the command task process, changes in environmental interference, and changes in operator interaction behavior. The existing image data enhancement methods fail to fully consider these complex and dynamic factors, lack deep integration with semantic information of the command scene, and thus the enhanced image data cannot accurately adapt to the actual needs of mobile command. SUMMARY

[0004] In view of the above-mentioned problems, in combination with the first aspect of the present application, the embodiments of the present application provide an image data enhancement method applied to a mobile command screen, the method comprising: obtaining the to-be-enhanced image data of the mobile command screen, real-time semantic data of the command scene, and enhancement effect feedback data, the to-be-enhanced image data containing command situation interactive image elements and instruction operation text image elements, the real-time semantic data of the command scene containing command task process change information, environmental interference change information, and operator interaction behavior information, and the enhancement effect feedback data containing corresponding records of enhancement operations and display effect evaluations in historical scenes; associating the image element types of the to-be-enhanced image data, the semantic dimensions of the real-time semantic data of the command scene, and the feedback dimensions of the enhancement effect feedback data, labeling semantic dependency relationships and influence degrees between different dimensions, and constructing command scene semantic association features; generating enhancement strategy self-adaption iteration rules based on the command scene semantic association features, the enhancement strategy self-adaption iteration rules containing logic for adjusting enhancement parameters according to semantic dependency relationships and logic for determining enhancement priorities according to influence degrees; calling a semantic perception enhancement model, loading the enhancement strategy self-adaption iteration rules, performing collaborative operations of situation interactive element enhancement and instruction text element enhancement on the to-be-enhanced image data, and generating preliminary enhanced image data; collecting feedback information after the preliminary enhanced image data is displayed on the mobile command screen, adjusting semantic dependency relationship and influence degree label of the semantic association features of the command scene according to the feedback information, updating the enhanced strategy self-adaptive iteration rule, performing secondary enhancement processing on the preliminary enhanced image data based on the updated enhanced strategy self-adaptive iteration rule, and generating final enhanced image data, wherein the feedback information includes visual attention behavior of the operator, command task execution progress, and screen display state change; transmitting the final enhanced image data to the display driving unit of the mobile command screen, storing the updated enhanced strategy self-adaptive iteration rule and the feedback information, and forming an associated record set.

[0005] In another aspect, the embodiment of the present application also provides an image data enhancement system applied to a mobile command screen, characterized in that it comprises: a processor, a machine readable storage medium for storing machine executable instructions of the processor, wherein the processor is configured to execute the above-mentioned image data enhancement method applied to the mobile command screen by executing the machine executable instructions.

[0006] In another aspect, the embodiment of the present application also provides a computer program product, which comprises machine executable instructions stored in a computer readable storage medium, and the processor of the image data enhancement system applied to the mobile command screen reads the machine executable instructions from the computer readable storage medium, and the processor executes the machine executable instructions, so that the image data enhancement system applied to the mobile command screen executes the above-mentioned image data enhancement method applied to the mobile command screen.

[0007] Based on the above aspects, by acquiring the to-be-enhanced image data of the mobile command screen, real-time semantic data of the command scene and enhancement effect feedback data, the dimensions of different types of data are associated and the semantic dependency relationship and influence degree are labeled, the command scene semantic association feature is constructed, the internal relationship between each data can be deeply mined, the enhancement strategy generated based on the special scene semantic association feature generates the adaptive iteration rule, the enhancement parameters and priority can be dynamically adjusted according to the semantic dependency relationship and influence degree, and the intelligentization and adaptability of the enhancement strategy are realized. The semantic perception enhancement model is called and the iteration rule is loaded, the to-be-enhanced image data is cooperatively operated to generate preliminary enhanced image data, and the quality and applicability of the image data are effectively improved. The feedback information of the preliminary enhanced image data is collected and the semantic association feature and the iteration rule are updated accordingly, and then secondary enhancement processing is performed to generate final enhanced image data, which can continuously adapt to the changes of the command scene and the user demand, and ensure that the finally output image data is always in the optimal state. The final enhanced image data is transmitted to the display driving unit and the related information is stored to form an association record set, further improving the efficiency and accuracy of image enhancement, and significantly improving the information display and interaction ability of the mobile command screen in the complex command scene. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 is the execution flow schematic diagram of the image data enhancement method applied to the mobile command screen provided by the embodiment of the application.

[0009] Figure 2 is the schematic diagram of exemplary hardware and software components of the image data enhancement system applied to the mobile command screen provided by the embodiment of the application. DETAILED DESCRIPTION

[0010] The application will be specifically described below in conjunction with the drawings of the specification, Figure 1 is the flow schematic diagram of the image data enhancement method applied to the mobile command screen provided by an embodiment of the application, and the image data enhancement method applied to the mobile command screen will be described in detail below.

[0011] Step S110: acquiring to-be-enhanced image data of the mobile command screen, real-time semantic data of the command scene and enhancement effect feedback data, the to-be-enhanced image data containing command situation interaction image elements and instruction operation text image elements, the real-time semantic data of the command scene containing command task progress change information, environmental interference change information and operator interaction behavior information, and the enhancement effect feedback data containing corresponding records of enhancement operations and display effect evaluation in historical scenes.

[0012] In the image data enhancement process of the mobile command screen, a multi-source data access mechanism needs to be established in the data acquisition link. The image data to be enhanced is received in real time from the command information system through high-definition video interfaces such as HDMI and DP. The data format supports common image formats such as RGB and YUV, and the frame rate is maintained above 30 fps to ensure the smoothness of dynamic pictures. The command situation interactive image elements are specific manifestations of image regions containing dynamic targets, such as moving icons in real-time refreshed resource allocation diagrams, and state-changing device operation monitoring graphics. The instruction operation text image elements are text regions in the screen for human-computer interaction, such as text labels on function buttons, option texts in drop-down menus, and prompt texts in input boxes. Real-time semantic data of the command scene is obtained through distributed data acquisition nodes. The command task progress change information covers the whole process state data from the start to the end of the task, including the time consumption of each stage and the proportion of completed work. Environmental interference change information is collected by sensor arrays deployed inside the command vehicle, including environmental light intensity values recorded by light sensors, vibration data collected by three-axis acceleration sensors, and interference signal strength captured by electromagnetic interference detectors. Operator interaction behavior information is collected by infrared touch sensors and binocular cameras integrated in the command screen frame, recording coordinate sequences, operation intervals, and gaze focus point trajectories. Enhancement effect feedback data is stored in a relational database, and each record contains enhancement parameter combinations, corresponding image frame IDs, clarity scores (1-5) and recognition scores (1-5) submitted by operators through a special evaluation interface, and the number of accidental touches recorded automatically by the system.

[0013] Step S120: associate the image element type of the image data to be enhanced, the semantic dimension of the real-time semantic data of the command scene, and the feedback dimension of the enhancement effect feedback data, label the semantic dependency relationship and influence degree between different dimensions, and construct the command scene semantic association feature.

[0014] Step S121: classify the image data to be enhanced, divide the command situation interactive image elements and the instruction operation text image elements, and extract the corresponding feature description information respectively. The feature description information of the command situation interactive image elements includes dynamic situation target type, situation change frequency, and situation display area. The feature description information of the instruction operation text image elements includes text instruction type, text display position, and text character density.

[0015] When classifying the augmented image data, an image classification model based on the Transformer architecture is used. The image classification model first divides the input image into fixed-size image blocks, converts each image block into a feature vector through linear projection, and adds position encoding to retain spatial position information. The feature vector sequence is input into a multi-head self-attention layer composed of multiple attention heads, which captures global feature correlations by calculating attention weights between different image blocks. The output of the self-attention layer is nonlinearly transformed by a feedforward neural network, and then stabilized by residual connection and layer normalization to stabilize the training process. After passing through multiple encoding blocks, the output feature vector is sent to the classification head, which is composed of a fully connected layer and a softmax function, and outputs the probability value of each image region belonging to the command situation interaction image element or the instruction operation text image element. According to the principle that the probability value is greater than a preset threshold (such as 0.5), pixel-level segmentation of the two types of elements is realized.

[0016] For the command situation interaction image element, the dynamic situation target type is determined by target detection algorithm, such as detecting dynamic targets in the image based on the YOLO algorithm, matching the preset target type library according to the contour features and motion parameters (such as speed and acceleration) of the target, and the target type library contains category labels such as ground transportation equipment and air monitoring equipment; The situation change frequency is calculated by counting the number of changes in the target state parameters (such as position, shape, and color) in the continuous N frames of images, and calculating the average number of changes per unit time; The situation display area is represented by the boundary box coordinates obtained by target detection, and stored as the left upper corner and right lower corner coordinate values in the image coordinate system, accurate to the pixel level.

[0017] For the instruction operation text image element, the text instruction type is classified by inputting the text content extracted by optical character recognition (OCR) technology into the BERT text classification model. After the BERT model performs word segmentation and word vector conversion on the text, it captures the context semantic features through the bidirectional Transformer encoder, and the [CLS] marker corresponding to the feature vector is input into the classification layer to output the probability distribution of the text belonging to types such as scheduling instructions, query instructions, and setting instructions. The type with the maximum probability is taken as the text instruction type; The text display position is determined by the text region boundary box coordinates obtained by OCR detection; The text character density is obtained by dividing the number of characters (obtained by OCR recognition result) in the text region by the area of the text region (calculated according to the boundary box coordinates), with the unit being characters per square pixel.

[0018] Step S122: Perform dimensional segmentation on the real-time semantic data of the command scenario into three dimensions: command task progress dimension, environmental interference dimension, and operator interaction dimension. Extract semantic description information for each dimension. The semantic description information for the command task progress dimension includes the current stage of the task, key nodes of the task, and changes in the urgency of the task. The semantic description information for the environmental interference dimension includes the trend of light intensity changes, the range of vibration impact, and the manifestation of electromagnetic interference. The semantic description information for the operator interaction dimension includes the type of operation action, the frequency of operation, and the area of ​​focus of operation.

[0019] When performing dimensional decomposition on real-time semantic data of command scenarios, an ontology-based semantic partitioning method is adopted. First, a semantic ontology model of the command scenario is constructed, defining three core semantic dimension classes: command task process class, environmental interference class, and operator interaction class, each containing several attributes. A semantic parser performs word segmentation, part-of-speech tagging, and named entity recognition on the raw semantic data, mapping the identified entities to the classes and attributes in the ontology model to achieve dimensional decomposition.

[0020] In the command and control task process dimension, the current stage of the task is determined by the stage identifier field in the task execution log. The stage identifier field is automatically generated by the command and control task management system according to the preset task flow, such as "initialization stage", "execution stage", "acceptance stage", etc. The key nodes of the task are determined by extracting the timestamp information from the task plan, such as the resource assembly completion time and data collection deadline. Each key node includes a node name and a planned completion time. The change in the urgency of the task is calculated by the ratio of the remaining time of the current task to the standard task duration. When the ratio is less than a preset threshold (such as 0.3), the urgency is determined to be increasing. Otherwise, it is determined to be stable or decreasing. The trend of change is represented by labels such as "increasing", "decreasing", and "stable".

[0021] In the environmental interference dimension, the light intensity change trend is determined by linearly fitting the continuous light intensity values ​​collected by the light sensor, calculating the slope of the fitted line. A positive slope indicates an increasing trend, while a negative slope indicates a decreasing trend. The absolute value of the slope represents the rate of change. The vibration influence range is determined by combining the vibration acceleration values ​​collected by the vibration sensor with the structural parameters of the command screen (such as screen size and material elasticity coefficient), using a vibration propagation model to calculate the distribution of screen displacement caused by vibration, defining the area where the displacement exceeds a preset threshold as the influence range, and representing it as a polygonal region in the screen coordinate system. The electromagnetic interference performance is determined by matching the spectrum characteristics of the interference signal collected by the electromagnetic interference detector with a preset interference type library (such as power frequency interference and radio frequency interference) to determine the interference type and interference intensity level (such as mild, moderate, and severe).

[0022] In the operator interaction dimension, the type of operation action is identified by matching the touch trajectory features (such as single-point click, multi-point zoom, and swipe direction) collected by the touch sensor with a preset action template library. The action template library contains trajectory feature parameters of various operation actions. The operation frequency is calculated by counting the number of effective operations per unit time (such as 1 minute). An effective operation refers to the operation that triggers the system response. The operation attention area is determined by clustering the coordinates of the gaze point collected by the eye-tracking system using a density clustering algorithm (such as DBSCAN). The image area where the cluster center is located is determined as the attention area and represented by the coordinates of a rectangular bounding box.

[0023] Step S123: Divide the enhanced effect feedback data into feedback dimensions: visual effect feedback dimension, task efficiency feedback dimension, and display stability feedback dimension. Extract feedback description information for each dimension. The feedback description information for the visual effect feedback dimension includes evaluation of element recognition clarity, color coordination, and detail rendering. The feedback description information for the task efficiency feedback dimension includes records of task operation time, number of operation errors, and task completion progress. The feedback description information for the display stability feedback dimension includes records of screen brightness changes, color deviation, and resolution maintenance.

[0024] When categorizing the feedback data for enhanced effects, the classification is based on the source of the feedback information and the evaluation objective. Visual effect feedback data comes from subjective scores submitted by operators through the evaluation interface and objective indicators from the image quality detection system; task efficiency feedback data comes from the operation logs of the command and task management system; and display stability feedback data comes from the real-time records of the screen hardware status monitoring module.

[0025] In the visual effect feedback dimension, the element recognition clarity evaluation adopts a combination of subjective scoring and objective indicators. The subjective scoring is the operator's score on the recognizability of image elements (1-10 points), and the objective indicator is the image clarity value calculated based on the structural similarity index (SSIM). The weighted average of the two is used as the final evaluation result. The color coordination evaluation analyzes the color histogram of each element in the image, calculates the hue difference and saturation difference of the main color, compares it with the preset coordination threshold, and generates evaluation labels such as coordinated, relatively coordinated, and uncoordinated. The detail rendering evaluation detects the clarity of preset detail feature points (such as texture edges and small marks) in the image and counts the proportion of clear feature points to the total feature points.

[0026] In the task efficiency feedback dimension, the task operation time is recorded as the time interval from when the operator starts the operation to when the operation is completed and confirmed by the system, accurate to milliseconds; the number of operation errors is recorded as the number of times the system error message is triggered during the operation, and the error message includes types such as invalid operation and parameter out of bounds; the task completion progress is recorded as the ratio of the number of currently completed task sub-items to the total number of sub-items, expressed as a percentage.

[0027] In the display stability feedback dimension, screen brightness change recording involves collecting continuous brightness values ​​from a brightness sensor, calculating the difference between the maximum and minimum brightness, and the standard deviation of brightness fluctuations; color deviation recording involves comparing the standard color chart image displayed on the screen with preset standard color values, calculating the deviation values ​​of the three RGB channels, in units of ΔE (calculated using the CIE1976 color difference formula); and resolution maintenance recording involves periodically capturing images of the screen display, analyzing the number of pixels and sharpness of the images, determining whether there is a reduction in resolution, and recording the timestamp and magnitude of the resolution change.

[0028] Step S124: Construct a feature node set, which includes image element nodes, semantic dimension nodes, and feedback dimension nodes. Image element nodes carry feature description information of the corresponding element, semantic dimension nodes carry semantic description information of the corresponding dimension, and feedback dimension nodes carry feedback description information of the corresponding dimension.

[0029] When constructing the feature node set, a node representation method from a graph database (such as Neo4j) is used. Each node contains a unique node ID, a node type label, and an attribute dictionary. The image element node type label is "Command and Control Interaction Image Element" or "Instruction and Operation Text Image Element," and the attribute dictionary stores the feature description information extracted in step S121, such as dynamic situation target type and text character density. The semantic dimension node type label is "Command and Control Task Progress Dimension," "Environmental Interference Dimension," or "Operator Interaction Dimension," and the attribute dictionary stores the semantic description information extracted in step S122, such as the current stage of the task and the trend of light intensity changes. The feedback dimension node type label is "Visual Effect Feedback Dimension," "Task Efficiency Feedback Dimension," or "Display Stability Feedback Dimension," and the attribute dictionary stores the feedback description information extracted in step S123, such as element recognition clarity evaluation and task operation time recording. Node IDs are generated using UUID format to ensure global uniqueness.

[0030] Step S125: Establish connections between nodes. Connections between image element nodes and semantic dimension nodes are constructed based on the relationship between element type and semantic dimension. Command situation interaction image element nodes are connected with command task progress dimension nodes and environmental interference dimension nodes. Command operation text image element nodes are connected with operator interaction dimension nodes and command task progress dimension nodes. Connections between image element nodes and feedback dimension nodes are constructed based on the relationship between element type and feedback dimension. Command situation interaction image element nodes are connected with visual effect feedback dimension nodes and task efficiency feedback dimension nodes. Command operation text image element nodes are connected with visual effect feedback dimension nodes and display stability feedback dimension nodes.

[0031] When establishing connections between nodes, association rules are determined based on domain knowledge and historical data statistical analysis. Between image element nodes and semantic dimension nodes, the frequency of semantic dimension changes affecting image element display requirements in historical augmentation cases is analyzed. When the frequency exceeds a preset threshold (e.g., 60%), an association edge is established. For example, changes in the command task progress dimension lead to adjustments in the display parameters of command posture interactive image elements in 80% of cases, thus establishing an association edge between them; the environmental interference dimension affects the display stability of command posture interactive image elements 75% of the time, establishing an association edge; the operator interaction dimension affects the interaction requirements of instruction operation text image elements 90% of the time, establishing an association edge; and the command task progress dimension affects the priority of instruction operation text image elements 70% of the time, establishing an association edge.

[0032] Between image element nodes and feedback dimension nodes, association edges are established based on the correlation analysis results between feedback information and image element types. The visual effect feedback dimension is related to the display quality of both types of image elements, with influence frequencies of 85% for command and control interaction image elements and 92% for instruction and operation text image elements; therefore, association edges are established for both. The task efficiency feedback dimension is mainly related to the information transmission efficiency of command and control interaction image elements, with an influence frequency of 88%, and an association edge is established for both. The display stability feedback dimension is mainly related to the continued readability of instruction and operation text image elements, with an influence frequency of 83%, and an association edge is established for both. In the graph database, association edges are represented as directed edges with direction, including edge ID, start node ID, end node ID, and association type label.

[0033] Step S126: Label the semantic dependencies of the associated edges and determine the dependency direction represented by each associated edge; label the degree of influence of the associated edges, and determine the degree of influence of each associated edge based on the records of the influence of different dimensions on the display effect of image elements in the historical enhancement effect feedback data. The degree of influence is divided according to the frequency of influence and the significance of the influence effect.

[0034] When labeling semantic dependencies of related edges, causal relationship detection algorithms (such as Granger causality tests) are used to analyze the time-series relationship between changes in semantic dimensions and changes in the display effect of image elements in historical data. If changes in semantic dimension A always precede changes in the display effect of image element B, and are statistically significantly correlated, then the dependency direction is determined to be A→B (semantic dimension node points to image element node). For example, if the "increased task urgency" event in the command task process dimension occurs, followed by the "increased dynamic cue intensity" event in the command situation interaction image element, and the Granger causality test shows a significant level (e.g., p<0.05), then the dependency direction of this related edge is labeled as the command task process dimension node pointing to the command situation interaction image element node.

[0035] When annotating the influence of associated edges, perform the following sub-steps: Step S1261: Extract the impact records of different semantic dimensions on the display effect of image elements in historical scenes from the enhancement effect feedback data. Each impact record includes the semantic dimension type, image element type, display effect change, and number of times the impact occurred.

[0036] The database containing the fields "semantic dimension change" and "display effect change" is queried. By associating the image element IDs, the semantic dimension type (such as command mission process dimension), image element type (such as command posture interactive image element), display effect change (such as clarity score change value), and number of times the effect occurs (total number of times the combination occurs) are extracted to form a structured effect record table.

[0037] Step S1262: Calculate the frequency of the influence of each semantic dimension and image element type combination from the influence record, and calculate the proportion of the number of times the combination appears in all historical scenes to the total number of scenes.

[0038] The impact record table is grouped according to the combination of "semantic dimension type - image element type". The number of impact occurrences in each group is counted, and then divided by the total number of historical scenarios (i.e., the number of scenarios in all feedback records in the database) to obtain the impact occurrence frequency. For example, the impact occurrence count of the combination of "command mission process dimension - command situation interaction image element" is 1200, the total number of scenarios is 1500, and the impact occurrence frequency is 1200 / 1500 = 0.8.

[0039] Step S1263: Analyze the changes in display effect corresponding to the combination of each semantic dimension and image element type, and determine the significance of the changes in display effect. If the change in semantic dimension causes the change in the evaluation index of the display effect of image elements to exceed the preset significant change range, it is marked as a significant impact; otherwise, it is marked as a non-significant impact.

[0040] The changes in evaluation metrics (such as sharpness score and recognizability score) in the display effect are compared with the preset significant change range (e.g., a change in sharpness score ≥ 1.5 points, with a maximum score of 5 points). If the change exceeds the range, it is considered a significant impact; otherwise, it is considered a non-significant impact. For example, in a certain impact record, the sharpness score changes from 3.0 to 4.8, a change of 1.8 points, which exceeds the significant change range of 1.5 points and is marked as a significant impact.

[0041] Step S1264: According to the set influence level classification standard, the association edges of all semantic dimensions and image element types are labeled with a level. The association edges between command situation interaction image element nodes and command task process dimension nodes, and the association edges between instruction operation text image element nodes and operator interaction dimension nodes are preferentially labeled as the first level. The first level corresponds to the combination where the influence occurs more than the preset frequency threshold and the display effect changes significantly. The second level corresponds to the combination where the influence occurs more than the preset frequency threshold but the display effect changes are not significant, or the influence occurs less than the preset frequency threshold but the display effect changes significantly. The third level corresponds to the combination where the influence occurs less than the preset frequency threshold and the display effect changes are not significant.

[0042] The preset frequency threshold is determined based on the frequency distribution of historical data, such as taking the 75th percentile of the frequency distribution (e.g., 0.75) as the threshold. For the combination of "command situation interaction image elements - command task progress dimension", if the frequency of influence is 0.8 > 0.75 and the display effect changes significantly, it is marked as Level 1; for the combination of "instruction operation text image elements - operator interaction dimension", if the frequency of influence is 0.9 > 0.75 and the display effect changes significantly, it is marked as Level 1; for the combination of "command situation interaction image elements - environmental interference dimension", if the frequency of influence is 0.75 (equal to the threshold) and the display effect changes significantly, it is marked as Level 1; for the combination of "instruction operation text image elements - command task progress dimension", if the frequency of influence is 0.7 < 0.75 but the display effect changes significantly, it is marked as Level 2; for the combination of "command situation interaction image elements - operator interaction dimension" (if it exists), if the frequency of influence is 0.6 < 0.75 and the display effect changes are not significant, it is marked as Level 3.

[0043] Step S1265: Extract the impact records of different feedback dimensions on the display effect of image elements in historical scenes from the enhancement effect feedback data. Each record includes the feedback dimension type, image element type, enhancement parameter adjustment status, and display effect optimization status.

[0044] Similar to step S1261, extract records of feedback dimension type (such as visual effect feedback dimension), image element type, enhancement parameter adjustment status (such as parameter adjustment amount), and display effect optimization status (such as score change after optimization).

[0045] Step S1266: Statistically analyze the frequency of the impact of each combination of feedback dimension and image element type, analyze the significance of display effect optimization, label the associated edges of the combination of feedback dimension and image element type, cross-validate the influence level of the labeled associated edges, compare the level labeling results of the same combination in different historical scenarios, if there are differences, re-analyze the impact records, adjust the level labeling until the results are consistent, bind the final determined influence level with the corresponding associated edge, record the basis of the level labeling, and form an associated edge influence level labeling document.

[0046] The frequency of the impact of the combination of statistical feedback dimension and image element type is analyzed to determine the significance of the optimized display effect (e.g., whether the score improvement after optimization exceeds a preset threshold). The data is then labeled according to the same grading standard as the semantic dimension. For example, the combination of "visual effect feedback dimension - instruction operation text image element" has an impact frequency of 0.95 > 0.75 and high optimization significance, and is labeled as level one. During cross-validation, historical data is divided into multiple subsets by time, and each subset is labeled with a level. If the consistency rate of the labeling results for each subset exceeds a preset threshold (e.g., 90%), the labeling is considered reliable; otherwise, outliers in the impact records are re-examined, and data collection errors or special interference scenarios are eliminated before re-labeling until the consistency rate reaches the target. The final impact level is stored as an attribute of the associated edges in the graph database, forming a labeling document containing the edge ID, level, and labeling basis (frequency value, significance judgment result).

[0047] Step S127: Organize all nodes, associated edges, semantic dependencies, and influence levels in a structured manner to generate semantic association features of the command scene. The semantic association features of the command scene include a node attribute table, an associated edge attribute table, a dependency table, and an influence level table.

[0048] Based on the graph database, structured tabular data is exported to form semantic association features of the command scene. The node attribute table includes node ID, node type, and all feature description information / semantic description information / feedback description information fields; the association edge attribute table includes edge ID, starting node ID, ending node ID, association type, and influence level; the dependency table includes edge ID, dependency direction (represented by "source node ID → target node ID"), and causality test results (such as p-value); the influence level table includes edge ID, frequency of influence occurrence, significance of display effect, and level labeling basis. The above tabular data is stored in JSON or CSV format as the basic data for subsequent generation of enhancement strategies.

[0049] Step S130: Generate an adaptive iteration rule for the enhancement strategy based on the semantic association features of the command scenario. The adaptive iteration rule for the enhancement strategy includes the logic of adjusting the enhancement parameters according to the semantic dependency relationship and the logic of determining the enhancement priority according to the degree of influence.

[0050] Step S131: Extract semantic dependencies from the semantic association features of the command scene, determine the semantic dimension nodes and feedback dimension nodes that affect each image element node, and form a dependency list of image elements and dimensions.

[0051] The dependency table is used to query all edges whose terminating node is an image element node. The starting nodes (semantic dimension nodes or feedback dimension nodes) of these edges are extracted and grouped by image element node ID to form a dependency list for each image element node. For example, the dependency list for image element node ID E1 in the command and control interaction includes: semantic dimension nodes D1 (command task progress dimension) and D2 (environmental interference dimension), and feedback dimension nodes F1 (visual effect feedback dimension) and F2 (task efficiency feedback dimension).

[0052] Step S132: For the interactive image elements of the command situation, determine the type of enhancement parameters that need to be adjusted based on the semantic dimension nodes in the dependency list of image elements and dimensions. Among them, the dynamic enhancement parameters of the command task process dimension affect the situation target, and the stability enhancement parameters of the environmental interference dimension affect the situation display.

[0053] Based on domain knowledge and parameter manuals, establish a mapping relationship between semantic dimension nodes and enhancement parameter types. Semantic descriptions of the command task process dimension (such as task urgency and key nodes) are related to dynamic enhancement parameters, including: dynamic target trajectory smoothness parameters (controlling the smoothness of the trajectory curve), dynamic cue intensity parameters (controlling the size / blinking frequency of the cue icon), and detail magnification ratio parameters (controlling the magnification factor of the target detail area). Semantic descriptions of environmental interference dimensions (such as light intensity and vibration) are related to stability enhancement parameters, including: brightness adaptive adjustment parameters (controlling the response speed and amplitude of brightness adjustment), color anti-interference compensation parameters (controlling the gain compensation value of the RGB channels), and jitter suppression intensity parameters (controlling the compensation coefficient of the image stabilization algorithm). Add these parameter types to the parameter adjustment list of the command situation interactive image elements.

[0054] Step S133: For the image elements of the instruction operation text, based on the semantic dimension nodes in the dependency list of image elements and dimensions, namely the operator interaction dimension and the command task process dimension, determine the type of enhancement parameters that need to be adjusted. The operator interaction dimension affects the interaction enhancement parameters of the text, and the command task process dimension affects the priority enhancement parameters of the text.

[0055] Similarly, a mapping is established between the semantic dimensions and parameter types of instruction operation text image elements. Semantic descriptions of operator interaction dimensions (such as operation type and area of ​​interest) are related to interaction enhancement parameters, including: text click response sensitivity parameters (controlling the click detection threshold), highlight contrast parameters (controlling the contrast between text and background), and zoom-in follow response speed parameters (controlling the response time of follow operations). Semantic descriptions of the command task progress dimension (such as task urgency) are related to priority enhancement parameters, including: text display hierarchy parameters (controlling the text's hierarchy in the display stack), transparency adjustment parameters (controlling the text's transparency), and highlight intensity parameters (controlling the display intensity of borders / shadows).

[0056] Step S134: Construct enhancement parameter adjustment logic. When the semantic description information of the semantic dimension node changes, adjust the value of the corresponding enhancement parameter according to the changed content. When the urgency of the task changes in the command task process dimension, increase the value of the dynamic enhancement parameter of the command situation interactive image element. When the light intensity change trend increases in the environmental interference dimension, increase the value of the stability enhancement parameter of the command situation interactive image element.

[0057] When building the enhanced parameter tuning logic, perform the following sub-steps: Step S1341: Perform change analysis on the semantic description information of the command task process dimension, extract the description content of the change in task urgency. If the description content indicates that the key node of the task is approaching or the remaining time limit for task execution is lower than the preset time limit threshold, the task urgency is determined to have increased. If the description content indicates that the task phase completion time is later than the planned time or the task resource pressure index is lower than the preset pressure threshold, the task urgency is determined to have decreased.

[0058] By analyzing the text description of changes in task urgency using natural language processing technology, keywords (such as "approaching critical node" and "insufficient remaining time") are extracted. Combined with the comparison between the remaining time limit and the preset time limit threshold (such as 20% of the total time), the direction of change in urgency is determined.

[0059] Step S1342: Determine the dynamic enhancement parameter type of the command situation interactive image elements, including motion trajectory enhancement parameters of situation targets, dynamic cue enhancement parameters of situation changes, and magnification enhancement parameters of situation details.

[0060] Clearly define the specific parameter names and physical meanings of the dynamic enhancement parameters. For example, the motion trajectory enhancement parameter is "trajectory smoothing factor" (value range 0-1), the dynamic prompt enhancement parameter is "prompt icon flashing frequency" (value range 1-5Hz), and the magnification enhancement parameter is "detail magnification factor" (value range 1-3 times).

[0061] Step S1343: Establish the correspondence between task urgency and dynamic enhancement parameters. When the task urgency increases, increase the value of motion trajectory enhancement parameters, increase the value of dynamic prompt enhancement parameters, and increase the value of magnification enhancement parameters.

[0062] Based on historical data, a functional relationship is fitted between the task urgency level (e.g., level 1-5) and parameter values. For example, for every level increase in urgency, the trajectory smoothing factor increases by 0.2, the flashing frequency increases by 1Hz, and the amplification factor increases by 0.5 times.

[0063] Step S1344: Perform change analysis on the semantic description information of the environmental interference dimension, extract the description content of the light intensity change trend. If the ambient light intensity value in the semantic description content is higher than the preset light intensity threshold, or the real-time fluctuation amplitude of the light intensity is higher than the preset fluctuation threshold, then it is determined that the light intensity change trend is enhanced; if the ambient light intensity value in the semantic description content is continuously in a stable range, or the real-time fluctuation amplitude of the light intensity is lower than the preset fluctuation threshold, then it is determined that the light intensity change trend is weakened.

[0064] Analyze the numerical description of the light intensity change trend, such as "light intensity rises rapidly" or "fluctuates violently", and determine the trend direction by comparing the light intensity value with a preset threshold (e.g., 500 lux) and the fluctuation amplitude with a preset fluctuation threshold (e.g., ±50 lux).

[0065] Step S1345: Determine the stability enhancement parameter type for the command situation interactive image elements, including brightness balance enhancement parameters for situation display, anti-interference enhancement parameters for situation color, and jitter suppression enhancement parameters for situation image.

[0066] Define the stability enhancement parameters, such as the brightness balance enhancement parameter being "brightness adjustment gain" (value range 0.5-2.0), the anti-interference enhancement parameter being "color compensation coefficient" (value range 0.8-1.2), and the jitter suppression enhancement parameter being "vibration compensation intensity" (value range 0-10).

[0067] Step S1346: Establish the correspondence between the light intensity change trend and the stability enhancement parameter. When the light intensity change trend is enhanced, increase the value of the brightness balance enhancement parameter, increase the value of the anti-interference enhancement parameter, and increase the value of the jitter suppression enhancement parameter.

[0068] The relationship between the fitted light intensity change trend level (e.g., level 1-5) and the parameter values ​​is shown. For example, for every 1 level increase in the trend level, the brightness adjustment gain increases by 0.3, the color compensation coefficient increases by 0.1, and the vibration compensation intensity increases by 2.

[0069] Step S1347: Perform change analysis on the semantic description information of the operator's interaction dimension, extract the description content of the operation action type, and if the operation frequency in the description content is higher than the preset frequency threshold, or the gaze duration on the instruction text area is higher than the preset duration threshold, it is determined that the operator's interaction demand for the instruction text has increased; if the operation focus in the description content shifts to the non-instruction area, or the duration of continuous no operation interval is higher than the preset interval threshold, it is determined that the interaction demand has decreased.

[0070] Analyze the descriptions of the operation actions (such as "high-frequency clicks" and "long-term gaze"), and combine the operation frequency with preset frequency thresholds (such as 5 times / minute) and gaze duration with preset duration thresholds (such as 3 seconds) to determine the changes in interaction requirements.

[0071] Step S1348: Determine the type of interactive enhancement parameters for the text image elements operated by the instruction, including text click response enhancement parameters, text highlighting enhancement parameters, and text zoom-in follow enhancement parameters.

[0072] Define the interaction enhancement parameters, such as the click response enhancement parameter being "click detection threshold" (value range 0.1-0.5N, analog quantity), the highlight enhancement parameter being "highlight contrast" (value range 1-3), and the zoom follow enhancement parameter being "follow response delay" (value range 0-200ms).

[0073] Step S1349: Establish the correspondence between interaction requirements and interaction enhancement parameters. When it is determined that the interaction requirement is enhanced, increase the value of the click response enhancement parameter, the highlight enhancement parameter, and the magnification follow enhancement parameter accordingly. When it is determined that the interaction requirement is weakened, decrease the value of the above enhancement parameters accordingly.

[0074] Establish a mapping between interaction requirement levels (levels 1-5) and parameter values. For example, if the requirement level increases by 1 level, the click detection threshold decreases by 0.1N (to improve sensitivity), the highlight contrast increases by 0.5, and the follow response latency decreases by 50ms.

[0075] Step S135: Extract the influence level of the associated edges in the semantic association features of the command scene, and determine the enhancement priority of the image elements in descending order of the level. The image elements corresponding to the first-level associated edges are given priority for enhancement operation. The adjustment priority of different enhancement parameters of the same image element is determined according to the influence level of the corresponding associated edges.

[0076] The highest influence level of the associated edges for each image element node is extracted from the associated edge attribute table, and the image element enhancement priority is determined by sorting them from highest to lowest level. For example, if the highest level of the associated edges for the command posture interaction image element is level one, and the highest level of the associated edges for the instruction operation text image element is level one (tied), then the number of associated edges at level one is further compared, with those having more being prioritized; if the number is the same, the average frequency of influence occurrence is compared, with those having higher values ​​being prioritized. The parameter adjustment priority for the same image element is sorted according to the influence level of its corresponding associated edges, with level one parameters taking precedence over level two parameters.

[0077] Step S136: Construct enhancement priority determination logic. In the command situation interaction image elements, dynamic enhancement parameters with higher correlation levels to the command task process dimension are adjusted first. In the instruction operation text image elements, interaction enhancement parameters with higher correlation levels to the operator interaction dimension are adjusted first.

[0078] For command and control situation interaction image elements, dynamic enhancement parameters correspond to the command task progress dimension (first level), and stability enhancement parameters correspond to the environmental interference dimension (first or second level). If the dynamic enhancement parameter has a higher correlation level (or a higher frequency of influence at the same level), then the dynamic enhancement parameter is adjusted first. For example, if the dynamic enhancement parameter has a correlation level of first level (frequency 0.8) and the stability enhancement parameter has a correlation level of first level (frequency 0.75), then the dynamic enhancement parameter takes priority. For command and control text image elements, interaction enhancement parameters correspond to the operator interaction dimension (first level), and priority enhancement parameters correspond to the command task progress dimension (second level). Therefore, interaction enhancement parameters are adjusted first.

[0079] Step S137: Set the enhancement strategy iteration trigger condition. When the change of semantic description information of real-time semantic data in the command scenario exceeds the preset semantic change range, or the feedback description information of the feedback information deviates from the preset feedback standard range, the enhancement strategy iteration update is triggered. Set the enhancement strategy iteration cycle. If the command scenario is in the critical node stage of the task, the iteration cycle is shortened. If the command scenario is in the stable stage of the task, the iteration cycle is extended.

[0080] The semantic change range is calculated by comparing the current value of the semantic description information with the value at the previous moment. The difference is calculated using cosine similarity or Euclidean distance. If the difference exceeds a preset threshold (e.g., 0.3), an iteration is triggered. Feedback deviating from the standard range refers to the evaluation index (e.g., clarity score) in the feedback description information being lower than a preset standard value (e.g., 3 points). The default iteration period is set to T. During the critical node phase of the task (e.g., 10 minutes before and after the critical node), the iteration period is shortened to T / 2, and during the stable phase of the task (e.g., non-critical nodes with stable urgency), the iteration period is extended to 2T.

[0081] Step S138: Integrate the enhancement parameter adjustment logic, enhancement priority determination logic, enhancement strategy iteration triggering condition, and enhancement strategy iteration cycle to form an enhancement strategy self-adaptive iteration rule. The enhancement strategy self-adaptive iteration rule includes a parameter adjustment comparison table, a priority sorting table, and an iteration control condition table for each image element.

[0082] The above logic and conditions are organized into a structured rule file. The parameter adjustment lookup table includes the semantic dimension change type, parameter adjustment direction, and adjustment amount calculation formula; the priority ranking table includes image element priority and parameter priority; and the iteration control condition table includes trigger thresholds and period settings. The rule file is stored in Extensible Markup Language (XML) format for easy model parsing and updating.

[0083] Step S140: Invoke the semantic perception enhancement model, load the self-adaptive iteration rules of the enhancement strategy, perform a collaborative operation of situational interaction element enhancement and instruction text element enhancement on the image data to be enhanced, and generate preliminary enhanced image data.

[0084] Step S141: Input the image data to be enhanced into the element recognition module of the semantic perception enhancement model, identify the display area of ​​the command posture interaction image element and the display area of ​​the instruction operation text image element, and generate element area positioning information.

[0085] The element recognition module receives the image data to be enhanced, performs element segmentation using the Transformer classification model trained in step S121, and outputs binary mask images of two types of elements. In the mask images, foreground pixels represent the corresponding element regions. Morphological processing (erosion, dilation) is performed on the mask images to remove noise. Then, the minimum bounding rectangle of the element region is extracted using a contour detection algorithm to obtain the coordinates of the top left corner (x1, y1) and bottom right corner (x2, y2) of the rectangle, generating element region positioning information, which is stored as a JSON array containing element ID, type, and coordinates.

[0086] Step S142: Load the enhancement priority determination logic in the self-adaptive iteration rule of the enhancement strategy. Based on the influence level of the associated edge, determine the image element that will be given priority for enhancement operation. If the associated edge level of the command situation interaction image element is higher than that of the instruction operation text image element, then the enhancement of the situation interaction element will be performed first.

[0087] The priority ranking table in the enhancement strategy iteration rule is parsed to obtain the priority ranking result of the current image element. For example, if the ranking result prioritizes the command situation interaction image element, the situation interaction element enhancement submodule is called first; if the command operation text image element prioritizes the command operation text image element, the command text element enhancement submodule is called first; if the priorities are the same, the two submodules are called in parallel (implemented through multithreading).

[0088] Step S143: Perform enhancement operations on the interactive image elements of the command situation, load the corresponding dynamic enhancement parameters and stability enhancement parameters in the enhancement parameter adjustment logic, adjust the motion trajectory display effect of the situation target according to the dynamic enhancement parameters, and adjust the brightness, color and jitter suppression effect of the situation display according to the stability enhancement parameters.

[0089] The values ​​of dynamic enhancement parameters and stability enhancement parameters under the current semantic dimension are retrieved from the enhancement parameter adjustment table. For the motion trajectory display effect, the original trajectory points are smoothed using a Bezier curve fitting algorithm based on the "trajectory smoothing factor" parameter value. The higher the smoothing factor, the greater the adjustment range of the curve control points. The dynamic prompt effect is achieved by controlling the periodic change of the icon's alpha channel to achieve blinking based on the "prompt icon blinking frequency" parameter value. The detail magnification effect is achieved by interpolating and magnifying the target area (such as bilinear interpolation) based on the "detail magnification factor" parameter value. The magnified area is then superimposed on the original position.

[0090] For stability enhancement, brightness adjustment is calculated based on the "brightness adjustment gain" parameter value and the current ambient light intensity value, using the formula "target brightness = base brightness × gain × light intensity compensation coefficient", which drives the backlight control circuit to adjust the brightness. For color anti-interference, the pixel values ​​of the RGB channels are multiplied by the compensation coefficient based on the "color compensation coefficient" parameter value, such as R'=R×Cr, G'=G×Cg, B'=B×Cb. For jitter suppression, the image displacement is calculated based on the "vibration compensation intensity" parameter value using an image stabilization algorithm (such as motion estimation based on feature point matching), and the pixels are compensated for inverse displacement. The compensation intensity parameter controls the weight of the compensation amount.

[0091] Step S144: During the enhancement of situational interaction elements, monitor the changes in real-time semantic data of the command scenario. If the change in semantic description information triggers the iteration condition, adjust the values ​​of dynamic enhancement parameters and stability enhancement parameters according to the self-adaptive iteration rules of the enhancement strategy.

[0092] The system receives semantic description information every second via a real-time data interface and compares the magnitude of change with the previously received information. If the magnitude of change exceeds the iteration trigger threshold (e.g., semantic change range of 0.3), the enhanced parameter adjustment lookup table is queried again to calculate the new target parameter value. The current parameter value is gradually adjusted to the target value using a parameter smoothing transition algorithm (e.g., exponential moving average). The transition time is set according to the parameter type (e.g., 0.5 seconds for dynamic parameters, 1 second for stable parameters) to avoid abrupt changes in the display effect.

[0093] Step S145: After completing the enhancement of the situational interaction elements, perform enhancement operations on the command operation text image elements, load the corresponding interaction enhancement parameters and priority enhancement parameters in the enhancement parameter adjustment logic, adjust the click response, highlighting and zoom-following effect of the text according to the interaction enhancement parameters, and adjust the display level and prominence of the text according to the priority enhancement parameters.

[0094] The query interaction enhancement parameters and priority enhancement parameters are set to specific values. The click response effect adjusts the touch sensor's signal detection threshold voltage based on the "Click Detection Threshold" parameter value; a lower threshold makes even slight touches easier to detect. The highlight display effect enhances contrast by adjusting the background brightness of the text area (reducing background brightness or increasing text brightness) based on the "Highlight Contrast" parameter value. The zoom-in follow effect, based on the "Follow Response Delay" parameter value, initiates a zoom-in operation after a specified delay after detecting the operator's gaze or touch point entering the text area; the zoom-in area is determined by the "Detail Magnification" parameter value.

[0095] In the priority enhancement effect, the display layer adjusts the Z-order value of the text drawing based on the "Text Display Layer" parameter value. The higher the value, the higher the priority for drawing on the top layer. The transparency adjusts the alpha value (0-1) of the text pixels based on the "Transparency Adjustment" parameter value. The lower the value, the less transparent. The highlighting effect controls the thickness of the text border and the blur radius of the shadow based on the "Highlighting Intensity" parameter value. The higher the intensity, the thicker the border and the more obvious the shadow.

[0096] Step S146: During the text element enhancement process, the scope of the text enhancement operation is controlled by combining the element area positioning information to ensure that it does not exceed the preset text display area.

[0097] Based on the text display area coordinates (x1, y1, x2, y2) in the element region positioning information, boundary constraints are set for the enhancement operation. For example, the magnification area boundary of the follow-up magnification effect must not exceed (x1-Δx, y1-Δy, x2+Δx, y2+Δy), where Δx and Δy are preset maximum expansion pixel values ​​(e.g., 50 pixels); the highlight display area is strictly limited to (x1, y1, x2, y2) to avoid affecting surrounding elements. Range control is achieved through a coordinate clipping algorithm, truncating or scaling enhancement effects that exceed the boundaries.

[0098] Step S147: Perform a collaborative calibration operation to enhance the command situation interaction image elements and the command operation text image elements. Based on the correlation between the two types of elements in the semantic association features of the command scene, adjust the display coordination between the situation interaction elements and the command text elements.

[0099] When performing a collaborative calibration operation, execute the following sub-steps: Step S1471: Extract the association relationship between command posture interaction image element nodes and instruction operation text image element nodes in the semantic association features of the command scene, and determine the display coordination requirements of the two types of elements. The display coordination requirements include color coordination requirements, position coordination requirements, and dynamic impact avoidance requirements.

[0100] Query the association relationship between two types of element nodes from the association edge attribute table of semantic association features in the command scene (if there is a direct association edge), or infer the indirect association relationship through the commonly associated semantic / feedback nodes, and determine the aspects that need to be coordinated: color (avoid color conflict), position (avoid area overlap), and dynamic influence (avoid dynamic elements interfering with static text).

[0101] Step S1472: To meet the color coordination requirements, analyze the current color display parameters of the two types of elements, extract the main color tone of the command posture interaction image element and the main color tone of the instruction operation text image element, calculate the hue difference value of the two main colors, and if the hue difference value exceeds the preset color coordination threshold, adjust the main color tone of one type of element.

[0102] The dominant hue (RGB values) of the two types of elements are extracted through color histogram analysis. After conversion to the HSV color space, the difference value ΔH of the hue (H) channel is calculated. The preset color coordination threshold is ΔH_threshold (e.g., 30°). If ΔH > ΔH_threshold, it needs to be adjusted.

[0103] Step S1473: When adjusting the color, based on the influence level of the associated edge recorded in the semantic association features of the command scene, select the main color of the element with a relatively low influence level of the associated edge to adjust, so that the adjusted hue difference value is less than or equal to the color coordination threshold.

[0104] Compare the influence levels of the associated edges between two types of elements. Elements with lower levels have more room for adjustment in their main color tone. For example, if the associated edge level of the command posture interactive image element is level one, and the associated edge level of the instruction operation text image element is also level one, then compare the frequency of their color influence on visual feedback, and prioritize adjusting the one with lower frequency. The adjustment method is to fine-tune the H value while maintaining the original saturation and brightness, so that ΔH ≤ ΔH_threshold.

[0105] Step S1474: In response to the position coordination requirements, based on the element area positioning information, detect the overlapping area of ​​the display areas of the two types of elements. If the overlapping area exceeds the preset overlap threshold, adjust the display position of one type of element in the preset direction away from the key information area of ​​the command task.

[0106] Calculate the area of ​​the intersection region of the bounding rectangles of the two types of elements. The ratio of this area to the area of ​​the smaller element is the overlap rate. If the overlap rate is greater than a preset threshold (e.g., 10%), adjustment is required. The key information area is determined based on the display area associated with the "key nodes of the task" in the command task process dimension. The adjustment direction is the direction away from this area (e.g., up, down, left, right, prioritizing the direction that reduces the overlap rate the most).

[0107] Step S1475: If the position adjustment causes the element display area to exceed the effective display boundary of the mobile command screen, then reduce the display ratio of one type of element according to the preset rules, and prioritize reduce the element with the lowest degree of influence of the associated edge.

[0108] Check if the bounding rectangle of the adjusted element exceeds the screen boundary (0≤x≤W, 0≤y≤H, where W and H are the screen resolution). If it does, calculate the scaling factor k=min((W-x1) / w, (H-y1) / h, x2 / w, y2 / h), where w=x2-x1 and h=y2-y1, and scale the element down proportionally. Select the element with the lowest influence level of its associated edges for scaling down; if the influence levels are the same, select the element with the larger area for scaling down.

[0109] Step S1476: In response to the need for dynamic impact avoidance, analyze the dynamic change range and frequency of the command situation interactive image elements. If the overlap area between the dynamic change range and the display area of ​​the command operation text image elements exceeds the preset dynamic overlap threshold, or the dynamic change frequency exceeds the preset visual comfort frequency threshold, then adjust the dynamic change range of the situation elements to maintain a minimum safe distance of not less than the preset minimum safe distance from the text display area.

[0110] The dynamic change range is determined by analyzing the maximum displacement area of ​​the situational element in consecutive frames, and calculating the overlap area between this area and the text area. The dynamic change frequency is the number of frames per second (FPS) of the situational element's state change. The preset dynamic overlap threshold is 50% of the static overlap threshold, and the visual comfort frequency threshold is 30 FPS. If the overlap area exceeds the limit or the frequency is >30 FPS, the dynamic change range is adjusted by restricting the element's movement boundary so that the minimum distance from the text area is ≥ a preset safety distance (e.g., 20 pixels).

[0111] Step S1477: If adjusting the dynamic change range will affect the integrity of the key information of the situation element, then reduce the dynamic change frequency of the situation element. The reduced frequency value shall not be lower than the preset minimum readable frequency threshold, and the reduction magnitude shall be positively correlated with the interaction demand intensity value of the text element.

[0112] The integrity of key information is checked using a target detection algorithm to ensure that the adjusted range includes all key targets (e.g., more than 90% of the targets). If not, the dynamic change frequency is reduced. The minimum readable frequency threshold is 15 FPS, and the reduction magnitude Δf = interaction demand intensity value × k (k is a proportional coefficient). The higher the interaction demand intensity value, the larger Δf is, ensuring that text reading is not affected by high-frequency dynamic interference.

[0113] Step S1478: After completing the adjustment of color, position, and dynamic effects, the real-time display screen of the mobile command screen is collected, and the visual attention behavior data of the operators is obtained through the visual acquisition device. If the duration of abnormal visual attention behavior exceeds the preset abnormal duration threshold, the collaborative calibration operation is repeated until the frequency of abnormal visual attention behavior is lower than the preset normal frequency threshold.

[0114] The system collects eye movement data from the operator via camera to detect abnormal visual attention behaviors (such as frequent blinking or wandering gaze). If the duration of the abnormality exceeds the abnormal duration threshold (e.g., 5 seconds), steps S1471 to S1477 are repeated. The maximum number of repetitions is a preset value (e.g., 3 times). If the target is still not met, a manual intervention prompt is issued.

[0115] Step S148: Integrate the enhanced image data into complete preliminary enhanced image data, record the parameter values, execution order, and collaborative calibration basis used in the enhancement process, and form a preliminary enhancement processing report.

[0116] The enhanced image data of the situation interaction elements and the enhanced image data of the command text elements are merged into the same image buffer based on the element region positioning information. An Alpha blending algorithm is used to process overlapping areas (if a small amount of overlap still exists). The image format is converted to a display format supported by the command screen (e.g., RGB888), and the resolution is adjusted to the screen's physical resolution. Enhancement parameter logs (parameter name, value, adjustment time), execution sequence logs (operation steps, start / end time), and collaborative calibration logs (adjustment type, before and after parameter comparison, calibration basis) are recorded, generating a preliminary enhancement processing report in PDF format.

[0117] Step S150: Collect feedback information after the preliminary enhanced image data is displayed on the mobile command screen. Adjust the semantic dependency and influence degree labeling of the semantic association features of the command scene according to the feedback information. Update the self-adaptive iteration rules of the enhancement strategy. Based on the updated self-adaptive iteration rules of the enhancement strategy, perform secondary enhancement processing on the preliminary enhanced image data to generate the final enhanced image data. The feedback information includes the operator's visual attention behavior, the progress of the command task execution, and changes in the screen display status.

[0118] Step S151: Collect the visual attention behavior of the operator through the visual acquisition device integrated in the mobile command screen. The visual acquisition device captures the operator's eye movements and gaze direction in real time, determines whether the operator's visual attention area is concentrated in the command posture interactive image element area or the instruction operation text image element area, and records the dwell time of the attention area and the switching frequency of the attention area.

[0119] The visual acquisition device uses an infrared eye-tracking camera with a sampling frequency of 30Hz. The coordinates (gx, gy) of the gaze's point on the screen are calculated using corneal reflection. These coordinates are compared with the element's location information to determine the area of ​​interest (situational area, text area, or other areas). The dwell time is the total time spent continuously in the same area, and the switching frequency is the number of times the gaze is switched between different areas per unit time.

[0120] Step S152: If the operator's visual attention area deviates from the command posture interaction image element area and the instruction operation text image element area for more than a preset time, or if the frequency of attention area switching exceeds the preset frequent switching range, it is marked as abnormal visual attention behavior. If the attention area is stable in one type of element area and the dwell time meets the preset effective attention range, it is marked as normal visual attention behavior.

[0121] The preset deviation time threshold is 5 seconds, and the frequent switching range is 5 times / minute. If the focus is on "other areas" for 5 consecutive seconds, or the switching frequency is >5 times / minute, it is marked as abnormal behavior; if the focus area is stable in the situation or text area, and the single dwell time is 2-30 seconds (effective focus range), it is marked as normal behavior.

[0122] Step S153: Collect the execution progress of the command task through the command task management system, record the completion stage of the command task, the number of completed operation steps, the number of incomplete operation steps, and the execution interval of operation steps during the preliminary enhanced image data display period, and determine whether the task execution progress meets the preset progress plan. If the progress is lagging behind, it is marked as an abnormal task execution; if the progress is normal, it is marked as a normal task execution.

[0123] Obtain task execution data from the command and task management system API interface, compare the completed stage with the stage nodes in the schedule plan, and calculate the percentage of progress as the number of completed steps / total number of steps. Compare this percentage with the planned percentage. If the progress percentage is less than the planned percentage - 5% (preset allowable deviation), it is marked as abnormal. If the interval between operation steps is greater than the average interval time × 1.5 times, it is also marked as abnormal.

[0124] Step S154: If the task execution progress is delayed, analyze whether the reason for the delay is related to the display effect of the preliminary enhanced image data, compare the correlation between the execution progress and the enhancement effect of the same task in the past, and if they are related, record the image element display problem corresponding to the delay.

[0125] The association strength (support and confidence) between "low display effect evaluation" and "delayed task progress" in historical data is analyzed using association rule mining algorithms (such as Apriori). If the confidence is greater than a preset threshold (such as 0.7), the association is determined. Specific display problems are recorded, such as "unobvious dynamic prompts for situational targets lead to information acquisition delays" and "excessive text character density increases reading time."

[0126] Step S155: Collect changes in screen display status through the hardware status monitoring unit of the mobile command screen, monitor the brightness changes, color changes, and resolution maintenance of different areas of the screen. If the brightness change of any area exceeds the preset stable range, or the color deviates, or the resolution fluctuates, it is marked as abnormal display status; otherwise, it is marked as normal display status.

[0127] The hardware status monitoring unit includes brightness sensors (sampling frequency 1Hz) and color sensors (sampling frequency 0.5Hz) distributed at the four corners and center of the screen. Resolution is read through the frame buffer (sampling frequency 0.1Hz). Brightness change amplitude = |current brightness - average brightness| / average brightness. An amplitude exceeding ±10% (stable range) is considered abnormal; color deviation ΔE > 5 (CIE1976) is considered abnormal; resolution fluctuation > ±1% is considered abnormal.

[0128] Step S156: Classify and organize the collected visual attention behavior data, count the number of occurrences and time periods of normal and abnormal behaviors, analyze the correlation between abnormal behaviors and element enhancement parameters in the preliminary enhanced image data, and determine the enhancement parameter types that may lead to abnormal behaviors; divide the command task execution progress data into stages, record the enhancement parameter values ​​of the preliminary enhanced image data for each task stage, analyze the correspondence between the progress and parameter values ​​of different stages, and identify key parameters that may affect the progress; perform time correlation on the screen display status change data, correlate the occurrence time of display status anomalies with the enhancement operation time of the preliminary enhanced image data, determine whether the display status anomaly occurred after the adjustment of a specific enhancement parameter, and if so, record the correlation between the enhancement parameter and the display status anomaly; integrate the classified and organized visual attention behavior data, the staged task execution progress data, and the time-correlated display status change data in the order of collection time to form a feedback information set where each data point has a time stamp, and the feedback information set includes feedback type, feedback description, associated elements, and associated parameter information.

[0129] Visual attention behavior data is categorized by timestamp to count the number of anomalies. Pearson correlation coefficient analysis is used to determine the correlation between the number of anomalies and the values ​​of enhancement parameters, identifying parameters with a correlation coefficient > 0.6 as suspicious parameters. Task progress data is divided into stages (e.g., every 10 minutes), recording the parameter values ​​for each stage. Linear regression analysis is used to analyze the relationship between parameter values ​​and progress percentages, identifying key parameters with significant impact (p < 0.05). Display status anomaly data is compared with enhancement operation timestamps; if the anomaly occurs within 5 seconds of the operation, it is considered relevant. Each record in the integrated feedback information set includes: feedback type (visual / task / display), feedback description (e.g., "anomaly dwell time 5 seconds"), associated element ID, associated parameter name and value range, with timestamps accurate to milliseconds.

[0130] Step S157: Adjust the semantic dependencies and influence levels of the semantic association features in the command scenario based on the feedback information, and update the adaptive iteration rules of the enhancement strategy, including the following sub-steps: Step S1571: Analyze the visual attention behavior data in the feedback information. If the abnormal visual attention behavior of the operator to the command situation interaction image elements is related to the semantic description information of the environmental interference dimension, then strengthen the semantic dependency relationship between the command situation interaction image element nodes and the environmental interference dimension nodes. If it is related to the semantic description information of the command task process dimension, then strengthen the semantic dependency relationship between the command situation interaction image element nodes and the command task process dimension nodes.

[0131] The causal relationship between visual attention to abnormal behavior and changes in the environmental interference dimension was analyzed using Granger causality tests. If the correlation was significant (p<0.05), the semantic dependency between the two was strengthened (e.g., by increasing the influence frequency weight of the associated edge). Similarly, the relationship with the command task progress dimension was analyzed and processed.

[0132] Step S1572: If the visual attention abnormal behavior is related to the instruction operation text image element and to the semantic description information of the operator interaction dimension, then strengthen the semantic dependency relationship between the instruction operation text image element node and the operator interaction dimension node. If it is related to the semantic description information of the task process dimension, then strengthen the semantic dependency relationship between the instruction operation text image element node and the task process dimension node.

[0133] Similar to step S1571, analyze the relationship between visual attention anomalies of text image elements operated by instructions and related semantic dimensions, and strengthen the dependency relationship (such as increasing the influence level of associated edges).

[0134] Step S1573: Adjust the influence level of the associated edges. If the feedback information shows that the enhancement effect of the element corresponding to a certain type of associated edge is not good, then reduce the influence level of the associated edge. If the enhancement effect is good, then increase the influence level of the associated edge.

[0135] If, after the element corresponding to a certain associated edge is enhanced, the rate of reduction of abnormal behavior in the feedback information is greater than a preset threshold (e.g., 30%), then its influence level is increased; if the rate of increase of abnormal behavior is greater than 10%, then the level is decreased.

[0136] Step S1574: For the progress lag in the command task execution progress data, if the lag is related to the dynamic enhancement parameters of the command situation interaction image elements, adjust the semantic dependency relationship corresponding to the dynamic enhancement parameters and increase the association between the task efficiency feedback dimension node and the situation element node. If the lag is related to the interaction enhancement parameters of the instruction text elements, increase the association between the task efficiency feedback dimension node and the text element node.

[0137] If the delay is related to the dynamic enhancement parameters (analysis results of step S156), then add a correlation edge of "task efficiency feedback dimension node → command situation interaction image element node" to the semantic association features of the command scene, and mark the dependency direction and initial influence level (such as the second level); similarly process the instruction text elements.

[0138] Step S1575: For display anomalies in the screen display status change data, if the anomaly is related to the stability enhancement parameters of the situation element, strengthen the semantic dependency between the situation element node and the environmental interference dimension node; if it is related to the display enhancement parameters of the text element, strengthen the semantic dependency between the text element node and the environmental interference dimension node.

[0139] Based on the correlation results between the displayed anomalies and the parameters, strengthen the corresponding semantic dependencies (such as increasing the causal weight of the associated edges).

[0140] Step S1576: Based on the adjusted semantic dependencies and influence levels, update the enhancement parameter adjustment logic in the self-adaptive iteration rules of the enhancement strategy, reset the correspondence between semantic dimension changes and enhancement parameter values, and ensure that the enhancement parameter adjustment magnitude for dimensions with increased associated edge levels exceeds the preset adjustment magnitude threshold, while the enhancement parameter adjustment magnitude for dimensions with decreased associated edge levels does not exceed the preset adjustment magnitude threshold.

[0141] For example, for a semantic dimension whose associated edge level is upgraded from level 2 to level 1, the parameter adjustment coefficient increases from 1.0 to 1.5 (the adjustment threshold is 1.2); for a dimension whose level is downgraded, the adjustment coefficient decreases from 1.0 to 0.8 (not exceeding the threshold).

[0142] Step S1577: Update the enhancement priority determination logic, reorder the enhancement priorities according to the adjusted impact level, enhance the image elements or parameters corresponding to the associated edges with higher levels first, and enhance the image elements or parameters corresponding to the associated edges with lower levels later.

[0143] Re-execute the priority sorting process in step S135 and update the priority sorting table.

[0144] Step S1578: Update the iteration triggering conditions of the enhancement strategy. If the frequency of abnormal feedback in the feedback information exceeds the preset abnormal frequency threshold, the abnormal feedback includes visual attention abnormality, task progress lag, and display abnormality. Then, reduce the semantic change amplitude threshold of the iteration trigger to improve the sensitivity of the strategy iteration. If the frequency of abnormal feedback is lower than the preset threshold, then increase the iteration trigger threshold to reduce unnecessary iterations.

[0145] The abnormal frequency threshold is set to 5 times / hour. If it exceeds this threshold, the semantic change amplitude threshold is reduced from 0.3 to 0.2. If it is less than 2 times / hour, the threshold is increased from 0.3 to 0.4.

[0146] Step S1579: Update the iteration cycle of the enhancement strategy. If the feedback information shows that the command scenario is in a critical stage of the mission, the iteration cycle is shortened based on the mission execution progress data. If it is in a stable stage of the mission, the iteration cycle is extended.

[0147] The critical phase of a task is determined by the current phase field (such as "approaching critical node"), and the iteration cycle is shortened from the default T to T / 2; the stable phase is extended to 2T.

[0148] Step S158: Based on the updated enhancement strategy's self-adaptive iteration rules, perform secondary enhancement processing on the initially enhanced image data to generate the final enhanced image data, and execute the following sub-steps: For example, step S1581: Load the updated enhancement strategy self-adaptive iteration rules, extract the enhancement parameter adjustment logic, enhancement priority determination logic, and iteration triggering conditions, and determine the parameter adjustment direction and execution order of the secondary enhancement process.

[0149] Parse the updated rule file to obtain the new parameter adjustment table, priority sorting table, and triggering conditions.

[0150] Step S1582: Compare the original enhancement parameters of the preliminary enhanced image data with the recommended parameters in the updated enhancement parameter adjustment logic to identify parameter differences. Parameter differences include differences in value, adjustment magnitude, and parameter type.

[0151] The original parameters are the parameter values ​​used during the initial enhancement, while the recommended parameters are the target values ​​obtained by querying and adjusting the comparison table based on the current semantic dimension. The difference value is calculated as: Recommended Parameters - Original Parameters. The adjustment magnitude difference is calculated as: Difference Value / Original Parameters. The parameter type difference refers to the type of parameters that are added or deleted.

[0152] Step S1583: For the interactive image elements of the command situation, if the value of the dynamic enhancement parameter in the updated recommended parameters is higher than the original parameter, then the motion trajectory enhancement effect and dynamic prompt enhancement effect of the situation target are improved. If the value of the stability enhancement parameter is higher than the original parameter, then the brightness balance, color anti-interference and jitter suppression effects of the situation display are adjusted.

[0153] The enhancement effect is adjusted according to the positive or negative direction of the parameter difference. For example, if the difference in the dynamic prompt enhancement parameter is positive, the flashing frequency is increased; if the difference in the stability parameter is positive, the brightness adjustment gain is increased.

[0154] Step S1584: If the value of one type of enhancement parameter in the updated recommended parameters is lower than that of the original parameter, the corresponding enhancement effect is reduced. The reduction range is determined according to the magnitude of the parameter difference; the larger the parameter difference, the greater the reduction range.

[0155] If the recommended value of the trajectory smoothing factor is lower than the original value, then the smoothing factor will be reduced to decrease the trajectory smoothness. The reduction magnitude is equal to the absolute value of the difference multiplied by the reduction coefficient.

[0156] Step S1585: For the text image element operated by the instruction, if the value of the interaction enhancement parameter in the updated recommended parameters is higher than the original parameter, then adjust the click response speed, highlighting effect and zoom-following effect of the text; if the value of the priority enhancement parameter is higher than the original parameter, then improve the display level and prominence of the text.

[0157] Increase the values ​​of interaction enhancement parameters, such as lowering the click detection threshold to improve response speed; increase priority parameters, such as increasing the display layer Z-order value.

[0158] Step S1586: If the value of the interaction enhancement parameter in the updated recommendation parameters is lower than that of the original parameter, the interaction enhancement effect of the text will be weakened. The degree of weakening will be determined based on the visual attention behavior data of the operator. The more records of normal visual attention behavior, the smaller the degree of weakening.

[0159] If the percentage of normal behavior records is >80%, the reduction coefficient is 0.5; if the percentage is <50%, the coefficient is 1.0.

[0160] Step S1587: During the secondary enhancement process, monitor the changes in feedback information in real time. If the newly collected feedback information is still abnormal, adjust the enhancement parameters again according to the updated iteration trigger conditions until the feedback information returns to normal.

[0161] After each parameter adjustment, wait for a preset observation time (e.g., 3 seconds) and collect new feedback information. If there are still abnormalities and the iteration condition is triggered, repeat steps S1582-S1586, for a maximum of 3 iterations.

[0162] Step S1588: Adapt the enhanced image data to the display characteristics of the mobile command screen, and adjust the overall display ratio and color output range of the image; after the adaptation is completed, generate the final enhanced image data, record the parameter adjustment content, feedback information monitoring results, and display adaptation basis during the secondary enhancement process, and form a secondary enhancement processing report.

[0163] Based on the display characteristics of the command screen (such as color gamut range sRGB / AdobeRGB, brightness range 200-500cd / m²), the color space of the enhanced image data is converted and the brightness range is compressed to ensure that the display effect meets the hardware specifications. A parameter adjustment log, feedback monitoring results (such as changes in the number of anomalies), and adaptation algorithm parameters (such as the transformation matrix) are recorded, and a secondary enhancement processing report is generated.

[0164] Step S160: Transmit the final enhanced image data to the display driver unit of the mobile command screen, and at the same time store the updated enhancement strategy self-adaptive iteration rules and feedback information to form a set of associated records.

[0165] The final enhanced image data is transmitted to the display driver board via the LVDS interface. The driver board converts the image data into panel control signals (such as TTL signals). The updated enhancement strategy iteration rules are stored in non-volatile memory (such as EEPROM), and the feedback information is stored in the database. The two are linked by timestamps and task IDs to form a set of associated records containing rule file paths, feedback information record IDs, task IDs, and timestamps.

[0166] Step S210: Train the semantic perception enhancement model. The semantic perception enhancement model includes an element recognition module, a parameter adjustment module, and a collaborative enhancement module. The training data includes labeled image data to be enhanced, corresponding real-time semantic data of the command scene, and enhancement effect feedback data.

[0167] Step S211: Construct the network structure of the semantic perception enhancement model. The element recognition module adopts a convolutional neural network structure, which includes an input layer, convolutional layer, pooling layer, fully connected layer and output layer; the parameter adjustment module adopts a recurrent neural network structure, which includes an input layer, hidden layer and output layer; the collaborative enhancement module adopts an attention mechanism network structure, which includes a feature input layer, attention calculation layer, feature fusion layer and output layer.

[0168] The element recognition module adopts the ResNet-50 architecture. The input layer receives a 224×224×3 image. The convolutional layer contains 5 convolutional stages (conv1-conv5), and each stage contains multiple convolutional blocks (including 3×3 convolution, batch normalization, and ReLU activation). The pooling layer uses 3×3 max pooling (stride 2). The fully connected layer contains 2 hidden layers (1024 and 512 neurons) and an output layer (2 neurons, corresponding to two types of elements).

[0169] The parameter adjustment module uses an LSTM network. The input layer receives a 128-dimensional semantic feature vector, the hidden layer contains 256 LSTM units (with dropout=0.3), and the output layer is a fully connected layer (outputting the amount of parameter adjustment, the dimension of which is equal to the number of parameters).

[0170] The collaborative enhancement module takes element features (256 dimensions) and parameter features (128 dimensions) as input. The attention calculation layer calculates attention weights (similarity between element features and parameter features) through a multilayer perceptron (MLP). The feature fusion layer performs a weighted summation of the element features (weights are attention values). The output layer outputs the parameter adjustment amount after collaborative enhancement.

[0171] Step S212: Collect large-scale training sample data, perform element type labeling and region labeling on the image data to be enhanced in the training sample data, perform dimension labeling on the real-time semantic data of the command scene, and perform evaluation index labeling on the enhancement effect feedback data.

[0172] Collect over 100,000 images of different command scenarios to be enhanced, and use the LabelMe tool to annotate element types and regions (polygon annotation); semantic data annotation dimension labels (e.g., "task urgency = high"); feedback data annotation evaluation index scores (1-5 points).

[0173] Step S213: Divide the labeled training sample data into a training set, a validation set, and a test set in a ratio of 7:2:1.

[0174] The sample order is randomly shuffled and divided into 70% training set, 20% validation set, and 10% test set to ensure that the category distribution of each set is consistent.

[0175] Step S214: Set the model training parameters. Set the initial learning rate to a preset value, the number of iterations to a preset number of rounds, the batch size to a preset number of samples, and the loss function to a combination of cross-entropy loss function and mean squared error loss function.

[0176] The initial learning rate is 0.001 (Adam optimizer), the number of iterations is 100, and the batch size is 32. The loss of the element recognition module is the cross-entropy loss (loss between class probability and label), the loss of the parameter adjustment module is the mean squared error loss (loss between predicted adjustment and actual adjustment), and the total loss is the weighted sum of the two (weight 1:1).

[0177] Step S215: Train the semantic awareness enhancement model using the training set, calculate the model output through forward propagation, update the model parameters through backpropagation, evaluate the model performance using the validation set in each iteration cycle, and stop training if the validation set loss does not decrease for several consecutive rounds.

[0178] The training process uses GPU acceleration (such as NVIDIA Tesla V100). After each training round, the loss value is calculated on the validation set. If the loss on the validation set does not decrease for 5 consecutive rounds, the learning rate is reduced to 0.1 times the original value, with a maximum of 3 reductions. If the loss still does not decrease, training is stopped and the optimal model parameters are saved.

[0179] Step S216: Use the test set to perform performance testing on the trained semantic perception enhancement model. The test metrics include element recognition accuracy, enhancement parameter adjustment error, and evaluation of collaborative enhancement effect. If the test metrics meet the preset standards, the model training is complete; otherwise, adjust the model structure or training parameters and retrain.

[0180] Element recognition accuracy = number of correctly recognized elements / total number of elements, target ≥ 95%; Parameter adjustment error = RMSE (predicted adjustment amount, actual adjustment amount), target ≤ 0.05; Collaborative enhancement effect evaluation = average subjective score, target ≥ 4.0 / 5.0. If the target is not met, adjust the number of network layers / neurons / learning rate and retrain.

[0181] In one exemplary embodiment, an image data enhancement system for a mobile command screen is provided. This system can be a terminal, server, etc., and its internal structure diagram can be as follows: Figure 2As shown, the image data enhancement system for a mobile command display includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computational and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements an image data enhancement method for a mobile command display. The display unit of the image data enhancement system applied to the mobile command screen is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the image data enhancement system applied to the mobile command screen can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the shell of the image data enhancement system applied to the mobile command screen, or external keyboards, touchpads, or mice, etc.

[0182] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. An image data enhancement method applied to a mobile command screen, characterized in that, The method includes: The system acquires the image data to be enhanced, real-time semantic data of the command scene, and enhancement effect feedback data of the mobile command screen. The image data to be enhanced includes command situation interaction image elements and command operation text image elements. The real-time semantic data of the command scene includes information on changes in the command task process, changes in environmental interference, and operator interaction behavior. The enhancement effect feedback data includes corresponding records of enhancement operations and display effect evaluations in historical scenes. The image element types of the image data to be enhanced, the semantic dimensions of the real-time semantic data of the command scene, and the feedback dimensions of the enhancement effect feedback data are associated. The semantic dependencies and influence levels between different dimensions are labeled to construct the semantic association features of the command scene. The enhancement strategy self-adaptive iteration rules are generated based on the semantic association features of the command scenario. The enhancement strategy self-adaptive iteration rules include the logic of adjusting the enhancement parameters according to the semantic dependency relationship and the logic of determining the enhancement priority according to the degree of influence. The semantic perception enhancement model is invoked, the self-adaptive iteration rules of the enhancement strategy are loaded, and the collaborative operation of situational interaction element enhancement and instruction text element enhancement is performed on the image data to be enhanced to generate preliminary enhanced image data. Feedback information is collected after the preliminary enhanced image data is displayed on the mobile command screen. Based on the feedback information, the semantic dependency relationship and influence degree labeling of the semantic association features of the command scene are adjusted, the self-adaptive iteration rules of the enhancement strategy are updated, and a second enhancement process is performed on the preliminary enhanced image data based on the updated self-adaptive iteration rules of the enhancement strategy to generate the final enhanced image data. The feedback information includes the operator's visual attention behavior, the progress of the command task execution, and changes in the screen display status. The final enhanced image data is transmitted to the display driver unit of the mobile command screen, while the updated enhancement strategy self-adaptive iteration rules and feedback information are stored to form a set of associated records.

2. The image data enhancement method applied to a mobile command screen according to claim 1, characterized in that, The process involves associating the image element types of the image data to be enhanced, the semantic dimensions of the real-time semantic data of the command scene, and the feedback dimensions of the enhancement effect feedback data, annotating the semantic dependencies and influence levels between different dimensions, and constructing semantic association features for the command scene, including: The image data to be enhanced is classified into elements, namely command and control interaction image elements and command and control text image elements, and the corresponding feature description information is extracted respectively. The feature description information of the command and control interaction image elements includes dynamic situation target type, situation change frequency, and situation display area. The feature description information of the command and control text image elements includes text command type, text display position, and text character density. The real-time semantic data of the command scenario is dimensionally split into three dimensions: command task process dimension, environmental interference dimension, and operator interaction dimension. Semantic description information is extracted for each dimension. The semantic description information of the command task process dimension includes the current stage of the task, key nodes of the task, and changes in the urgency of the task. The semantic description information of the environmental interference dimension includes the trend of light intensity change, the range of vibration influence, and the manifestation of electromagnetic interference. The semantic description information of the operator interaction dimension includes the type of operation action, operation frequency, and operation focus area. The feedback data for the enhanced effect is divided into three dimensions: visual effect feedback, task efficiency feedback, and display stability feedback. Feedback description information is extracted for each dimension. The feedback description information for the visual effect feedback dimension includes evaluations of element recognition clarity, color coordination, and detail rendering. The feedback description information for the task efficiency feedback dimension includes records of task operation time, number of operation errors, and task completion progress. The feedback description information for the display stability feedback dimension includes records of screen brightness changes, color deviation, and resolution retention. Construct a feature node set, which includes image element nodes, semantic dimension nodes, and feedback dimension nodes. Image element nodes carry feature description information of the corresponding element, semantic dimension nodes carry semantic description information of the corresponding dimension, and feedback dimension nodes carry feedback description information of the corresponding dimension. Establish connections between nodes. The connections between image element nodes and semantic dimension nodes are constructed based on the relationship between element type and semantic dimension. Command situation interaction image element nodes are connected with command task progress dimension nodes and environmental interference dimension nodes. Command operation text image element nodes are connected with operator interaction dimension nodes and command task progress dimension nodes. The connections between image element nodes and feedback dimension nodes are constructed based on the relationship between element type and feedback dimension. Command situation interaction image element nodes are connected with visual effect feedback dimension nodes and task efficiency feedback dimension nodes. Command operation text image element nodes are connected with visual effect feedback dimension nodes and display stability feedback dimension nodes. The semantic dependencies of associated edges are labeled to determine the dependency direction represented by each associated edge; the degree of influence of associated edges is labeled, and the degree of influence of each associated edge is determined based on the impact records of different dimensions on the display effect of image elements in the historical enhancement effect feedback data. The degree of influence is divided according to the frequency of influence and the significance of the effect. All nodes, associated edges, semantic dependencies, and influence levels are structured and organized to generate semantic association features for the command scenario. These semantic association features include a node attribute table, an associated edge attribute table, a dependency table, and an influence level table.

3. The image data enhancement method applied to a mobile command screen according to claim 2, characterized in that, The degree of influence of the labeled associated edges is determined based on the impact records of different dimensions on the display effect of image elements in historical enhancement effect feedback data, to determine the degree of influence of each associated edge, including: The impact records of different semantic dimensions on the display effect of image elements in historical scenarios are extracted from the enhancement effect feedback data. Each impact record includes the semantic dimension type, image element type, display effect change, and number of times the impact occurred. The frequency of the impact of each semantic dimension and image element type combination is statistically analyzed from the impact records, and the proportion of the number of times the combination appears in all historical scenes to the total number of scenes is calculated. Analyze the changes in display effect corresponding to each combination of semantic dimension and image element type, and determine the significance of the changes in display effect. If the change in semantic dimension causes the evaluation index of the display effect of image element to change beyond the preset significant change range, it is marked as a significant effect; otherwise, it is marked as a non-significant effect. According to the established impact level classification standard, the association edges of all semantic dimensions and image element types are labeled with a level. The association edges between command posture interaction image element nodes and command task process dimension nodes, and the association edges between instruction operation text image element nodes and operator interaction dimension nodes are preferentially labeled as the first level. The first level corresponds to the combination where the impact occurs more frequently than the preset frequency threshold and the display effect changes significantly. The second level corresponds to the combination where the impact occurs more frequently than the preset frequency threshold but the display effect changes are not significant, or the impact occurs less frequently than the preset frequency threshold but the display effect changes significantly. The third level corresponds to the combination where the impact occurs less frequently than the preset frequency threshold and the display effect changes are not significant. Extract the impact records of different feedback dimensions on the display effect of image elements in historical scenarios from the enhancement effect feedback data. Each record includes the feedback dimension type, image element type, enhancement parameter adjustment status, and display effect optimization status. The frequency of the impact of each combination of feedback dimension and image element type is statistically analyzed, the significance of display effect optimization is analyzed, the associated edges of the combination of feedback dimension and image element type are graded, the influence level of the graded associated edges is cross-validated, the graded labeling results of the same combination in different historical scenarios are compared, if there are differences, the impact records are re-analyzed, the graded labels are adjusted until the results are consistent, the final determined influence level is bound to the corresponding associated edge, the basis of the graded labeling is recorded, and an associated edge influence level labeling document is formed.

4. The image data enhancement method applied to a mobile command screen according to claim 1, characterized in that, The self-adaptive iterative rules for generating enhanced strategies based on semantic association features of command scenarios include: Extract semantic dependencies from the semantic association features of the command scene, determine the semantic dimension nodes and feedback dimension nodes that affect each image element node, and form a dependency list of image elements and dimensions. For the interactive image elements of the command situation, the types of enhancement parameters that need to be adjusted are determined based on the semantic dimension nodes in the dependency list of image elements and dimensions. Among them, the command task process dimension affects the dynamic enhancement parameters of the situation target, and the environmental interference dimension affects the stability enhancement parameters of the situation display. For the image elements of the instruction operation text, based on the semantic dimension nodes in the dependency list of image elements and dimensions, namely the operator interaction dimension and the command task process dimension, determine the type of enhancement parameters that need to be adjusted. The operator interaction dimension affects the interaction enhancement parameters of the text, and the command task process dimension affects the priority enhancement parameters of the text. Construct an enhancement parameter adjustment logic. When the semantic description information of the semantic dimension node changes, adjust the value of the corresponding enhancement parameter according to the changed content. When the urgency of the task changes in the command task process dimension, increase the dynamic enhancement parameter value of the command situation interactive image element. When the light intensity change trend increases in the environmental interference dimension, increase the stability enhancement parameter value of the command situation interactive image element. Extract the influence level of the associated edges in the semantic association features of the command scene, and determine the enhancement priority of the image elements in order of the level from high to low. The image elements corresponding to the first-level associated edges are given priority for enhancement operation. The adjustment priority of different enhancement parameters of the same image element is determined according to the influence level of the corresponding associated edges. Construct an enhancement priority determination logic: in the command situation interaction image elements, the dynamic enhancement parameters with higher correlation levels with the command task process dimension are adjusted first; in the instruction operation text image elements, the interaction enhancement parameters with higher correlation levels with the operator interaction dimension are adjusted first. Set the enhancement strategy iteration trigger conditions. When the change of semantic description information of real-time semantic data in the command scenario exceeds the preset semantic change range, or the feedback description information of the feedback information deviates from the preset feedback standard range, the enhancement strategy iteration update is triggered. Set an iteration cycle for the enhancement strategy. If the command scenario is at a critical stage of the mission, the iteration cycle will be shortened; if the command scenario is at a stable stage of the mission, the iteration cycle will be extended. The logic for enhancing parameter adjustment, the logic for determining enhancement priority, the triggering conditions for enhancement strategy iteration, and the iteration cycle of enhancement strategy are integrated to form the self-adaptive iteration rules for enhancement strategy. The self-adaptive iteration rules for enhancement strategy include a parameter adjustment reference table, a priority sorting table, and an iteration control condition table for each image element.

5. The image data enhancement method applied to a mobile command screen according to claim 4, characterized in that, When the semantic description information of the semantic dimension node changes, the values ​​of the corresponding enhancement parameters are adjusted according to the changes. When the urgency of the task changes in the command task process dimension, the dynamic enhancement parameter values ​​of the command situation interactive image elements are increased. When the light intensity change trend increases in the environmental interference dimension, the stability enhancement parameter values ​​of the command situation interactive image elements are increased, including: The semantic description information of the command task process dimension is analyzed for changes, and the description content of the changes in the urgency of the task is extracted. If the description content indicates that the key node of the task is approaching or the remaining time limit of the task execution is lower than the preset time limit threshold, the urgency of the task is determined to have increased. If the description content indicates that the completion time of the task stage is later than the planned time or the task resource pressure index is lower than the preset pressure threshold, the urgency of the task is determined to have decreased. Determine the types of dynamic enhancement parameters for interactive image elements of the command situation, including motion trajectory enhancement parameters for situational targets, dynamic cue enhancement parameters for situational changes, and magnification enhancement parameters for situational details; Establish a correspondence between task urgency and dynamic enhancement parameters. When task urgency increases, increase the value of motion trajectory enhancement parameters, increase the value of dynamic prompt enhancement parameters, and increase the value of magnification enhancement parameters. The semantic description information of the environmental interference dimension is analyzed for changes, and the description content of the light intensity change trend is extracted. If the ambient light intensity value in the semantic description content is higher than the preset light intensity threshold, or the real-time fluctuation amplitude of the light intensity is higher than the preset fluctuation threshold, the light intensity change trend is determined to be enhanced. If the ambient light intensity value in the semantic description content is continuously in a stable range, or the real-time fluctuation amplitude of the light intensity is lower than the preset fluctuation threshold, the light intensity change trend is determined to be weakened. Determine the types of stability enhancement parameters for interactive image elements of the command situation, including brightness balance enhancement parameters for situation display, anti-interference enhancement parameters for situation color, and jitter suppression enhancement parameters for situation image. Establish the correspondence between the light intensity change trend and the stability enhancement parameter. When the light intensity change trend is enhanced, increase the value of the brightness balance enhancement parameter, increase the value of the anti-interference enhancement parameter, and increase the value of the jitter suppression enhancement parameter. The semantic description information of the operator's interaction dimension is analyzed for changes. The description content of the operation action type is extracted. If the operation frequency in the description content is higher than the preset frequency threshold, or the gaze duration on the instruction text area is higher than the preset duration threshold, it is determined that the operator's need for interaction with the instruction text is enhanced. If the operation focus in the description content shifts to the non-instruction area, or the duration of continuous no operation interval is higher than the preset interval threshold, it is determined that the need for interaction is weakened. Determine the type of interactive enhancement parameters for the text image elements operated by the command, including text click response enhancement parameters, text highlighting enhancement parameters, and text zoom-in follow enhancement parameters; Establish a correspondence between interaction requirements and interaction enhancement parameters. When it is determined that the interaction requirement is enhanced, increase the values ​​of the click response enhancement parameter, the highlight enhancement parameter, and the zoom follow enhancement parameter accordingly. When it is determined that the interaction requirement is weakened, decrease the values ​​of the above enhancement parameters accordingly.

6. The image data enhancement method applied to a mobile command screen according to claim 1, characterized in that, The process involves invoking the semantic awareness enhancement model, loading the adaptive iteration rules of the enhancement strategy, and performing a collaborative operation of situational interaction element enhancement and instruction text element enhancement on the image data to be enhanced, generating preliminary enhanced image data, including: The image data to be enhanced is input into the element recognition module of the semantic perception enhancement model, which identifies the display area of ​​the command situation interaction image element and the display area of ​​the instruction operation text image element, and generates element area positioning information. The enhancement priority determination logic in the self-adaptive iteration rule of the loading enhancement strategy determines the image elements that will be enhanced first based on the influence level of the associated edges. If the associated edge level of the command situation interaction image element is higher than that of the instruction operation text image element, the enhancement of the situation interaction element will be performed first. Enhancement operations are performed on the interactive image elements of the command situation, loading the corresponding dynamic enhancement parameters and stability enhancement parameters in the enhancement parameter adjustment logic, adjusting the display effect of the motion trajectory of the situation target according to the dynamic enhancement parameters, and adjusting the brightness, color and jitter suppression effect of the situation display according to the stability enhancement parameters. During the enhancement of situational interaction elements, the changes in real-time semantic data of the command scenario are monitored in real time. If the change in semantic description information triggers the iteration condition, the values ​​of dynamic enhancement parameters and stability enhancement parameters are adjusted according to the self-adaptive iteration rules of the enhancement strategy. After enhancing the interactive elements of the situation, enhance the text and image elements of the command operation, load the corresponding interactive enhancement parameters and priority enhancement parameters in the enhancement parameter adjustment logic, adjust the click response, highlighting and zoom-following effect of the text according to the interactive enhancement parameters, and adjust the display level and prominence of the text according to the priority enhancement parameters. During the process of enhancing the command text element, the scope of the text enhancement operation is controlled by combining the element area positioning information to ensure that it does not exceed the preset text display area; The system performs a collaborative calibration operation to enhance the interactive image elements of the command situation and the image elements of the command operation text. Based on the correlation between the two types of elements in the semantic association features of the command scene, the system adjusts the display coordination between the interactive image elements and the command text elements. The enhanced image data is integrated into complete preliminary enhanced image data, and the parameter values, execution order, and collaborative calibration basis used in the enhancement process are recorded to form a preliminary enhancement processing report.

7. The image data enhancement method applied to a mobile command screen according to claim 6, characterized in that, The collaborative calibration operation of enhancing the interactive image elements of the command situation and enhancing the image elements of the command operation text adjusts the display coordination between the interactive image elements and the command text elements based on the association relationship between the two types of elements in the semantic association features of the command scene, including: Extract the association relationship between command situation interaction image element nodes and instruction operation text image element nodes in the semantic association features of the command scene, and determine the display coordination requirements of the two types of elements. The display coordination requirements include color coordination requirements, position coordination requirements, and dynamic impact avoidance requirements. To address color coordination requirements, the current color display parameters of the two types of elements are analyzed, the main color tone of the command posture interaction image elements and the main color tone of the instruction operation text image elements are extracted, the hue difference value of the two main color tones is calculated, and if the hue difference value exceeds the preset color coordination threshold, the main color tone of one type of element is adjusted. When adjusting colors, based on the influence level of the associated edges recorded in the semantic association features of the command scene, the main color of the element with a relatively low influence level of the associated edges is selected for adjustment, so that the adjusted hue difference value is less than or equal to the color coordination threshold. To address the need for location coordination, based on the element region positioning information, the overlapping area of ​​the display areas of two types of elements is detected. If the overlapping area exceeds a preset overlap threshold, the display position of one type of element is adjusted, and the adjustment direction is a preset direction that moves away from the key information area of ​​the command task. If the position adjustment causes the element display area to exceed the effective display boundary of the mobile command screen, the display ratio of one type of element will be reduced according to the preset rules, prioritizing the reduction of the element with the lowest degree of influence from the associated edge; To address the need for dynamic impact avoidance, the dynamic change range and frequency of command situation interactive image elements are analyzed. If the overlap area between the dynamic change range and the display area of ​​the command operation text image elements exceeds the preset dynamic overlap threshold, or the dynamic change frequency exceeds the preset visual comfort frequency threshold, the dynamic change range of the situation elements is adjusted to maintain a minimum safe distance of not less than the preset minimum safe distance from the text display area. If adjusting the dynamic change range affects the integrity of key information of the situation element, the dynamic change frequency of the situation element shall be reduced. The reduced frequency value shall not be lower than the preset minimum readable frequency threshold, and the reduction magnitude shall be positively correlated with the interaction demand intensity value of the text element. After adjusting the color, position, and dynamic effects, the real-time display screen of the mobile command screen is collected, and the visual attention behavior data of the operators is obtained through the visual acquisition device. If the duration of abnormal visual attention behavior exceeds the preset abnormal duration threshold, the collaborative calibration operation is repeated until the frequency of abnormal visual attention behavior is lower than the preset normal frequency threshold.

8. The image data enhancement method applied to a mobile command screen according to claim 1, characterized in that, The feedback information after the acquired preliminary enhanced image data is displayed on the mobile command screen includes: The visual attention behavior of operators is collected by the visual acquisition device integrated in the mobile command screen. The visual acquisition device captures the eye movements and gaze direction of the operators in real time, determines whether the operator's visual attention area is concentrated in the command situation interaction image element area or the instruction operation text image element area, and records the dwell time of the attention area and the switching frequency of the attention area. If the operator's visual attention area deviates from the command posture interactive image element area and the instruction operation text image element area for more than a preset time, or if the frequency of switching of the attention area exceeds the preset frequent switching range, it will be marked as abnormal visual attention behavior. If the attention area is stable in one type of element area and the dwell time is in line with the preset effective attention range, it will be marked as normal visual attention behavior. The command and control task management system collects the progress of command and control task execution, records the completion stage of the command and control task, the number of completed operation steps, the number of incomplete operation steps, and the execution interval of operation steps during the initial enhanced image data display period, and judges whether the task execution progress meets the preset progress plan. If the progress is lagging behind, it is marked as an abnormal task execution; if the progress is normal, it is marked as a normal task execution. If the task execution progress is delayed, analyze whether the reason for the delay is related to the display effect of the initial enhanced image data, compare the correlation between the execution progress and the enhancement effect of the same task in the past, and if there is a correlation, record the image element display problem corresponding to the delay. The hardware status monitoring unit of the mobile command screen collects changes in the screen display status, monitors the brightness changes, color changes, and resolution maintenance of different areas of the screen. If the brightness change of any area exceeds the preset stable range, or the color deviates, or the resolution fluctuates, it is marked as an abnormal display status; otherwise, it is marked as a normal display status. The collected visual attention behavior data are classified and organized, the frequency and time of occurrence of normal and abnormal behaviors are counted, the correlation between abnormal behaviors and element enhancement parameters in the preliminary enhanced image data is analyzed, and the types of enhancement parameters that may lead to abnormal behaviors are determined. The data on the progress of command and control tasks are divided into stages. The values ​​of enhancement parameters for the initial enhanced image data are recorded for each task stage. The relationship between the progress and parameter values ​​at different stages is analyzed to identify key parameters that may affect the progress. Time correlation is performed on the screen display status change data to correspond the occurrence time of display status abnormality with the enhancement operation time of the initial enhanced image data, and to determine whether the display status abnormality occurred after the adjustment of a specific enhancement parameter. If there is a correspondence, the correlation between the enhancement parameter and the display status abnormality is recorded. The categorized visual attention behavior data, the phased task execution progress data, and the time-linked display status change data are integrated in the order of collection time to form a feedback information set in which each data point is time-stamped. The feedback information set includes feedback type, feedback description, associated elements, and associated parameter information.

9. The image data enhancement method applied to a mobile command screen according to claim 1, characterized in that, The step of adjusting the semantic dependencies and influence levels of the semantic association features of the command scenario based on feedback information, and updating the adaptive iteration rules of the enhancement strategy, includes: Analyze the visual attention behavior data in the feedback information. If the abnormal visual attention behavior of the operator to the command situation interactive image elements is related to the semantic description information of the environmental interference dimension, then strengthen the semantic dependency relationship between the command situation interactive image element nodes and the environmental interference dimension nodes. If it is related to the semantic description information of the command task process dimension, then strengthen the semantic dependency relationship between the command situation interactive image element nodes and the command task process dimension nodes. If the visual attention to abnormal behavior is related to the instruction operation text image element and the semantic description information of the operator interaction dimension, then the semantic dependency relationship between the instruction operation text image element node and the operator interaction dimension node is strengthened. If it is related to the semantic description information of the task process dimension, then the semantic dependency relationship between the instruction operation text image element node and the task process dimension node is strengthened. Adjust the influence level of associated edges. If the feedback information shows that the enhancement effect of the element corresponding to a certain type of associated edge is not good, then reduce the influence level of the associated edge; if the enhancement effect is good, then increase the influence level of the associated edge. For the progress lag in the command mission execution progress data, if the lag is related to the dynamic enhancement parameters of the command situation interaction image elements, then adjust the semantic dependency relationship corresponding to the dynamic enhancement parameters and increase the association between the task efficiency feedback dimension node and the situation element node. If the lag is related to the interaction enhancement parameters of the command text elements, then increase the association between the task efficiency feedback dimension node and the text element node. For display anomalies in screen display status change data, if the anomaly is related to the stability enhancement parameters of the situation element, then strengthen the semantic dependency between the situation element node and the environmental interference dimension node; if it is related to the display enhancement parameters of the text element, then strengthen the semantic dependency between the text element node and the environmental interference dimension node. Based on the adjusted semantic dependencies and the level of influence, update the enhancement parameter adjustment logic in the self-adaptive iteration rules of the enhancement strategy, reset the correspondence between semantic dimension changes and enhancement parameter values, and ensure that the enhancement parameter adjustment magnitude for dimensions with increased associated edge levels exceeds the preset adjustment magnitude threshold, while the enhancement parameter adjustment magnitude for dimensions with decreased associated edge levels does not exceed the preset adjustment magnitude threshold. The enhancement priority determination logic has been updated. The enhancement priorities have been reordered according to the adjusted impact level. Image elements or parameters corresponding to associated edges with higher levels are enhanced first, while image elements or parameters corresponding to associated edges with lower levels are enhanced later. Update and enhance the triggering conditions for strategy iteration. If the frequency of abnormal feedback in the feedback information exceeds the preset abnormal frequency threshold, the abnormal feedback includes visual attention abnormalities, task progress delays, and display abnormalities. Then, reduce the semantic change amplitude threshold for iteration triggering to improve the sensitivity of strategy iteration. If the frequency of abnormal feedback is lower than the preset threshold, increase the iteration triggering threshold to reduce unnecessary iterations. The update and enhancement strategy iteration cycle is adjusted. If the feedback information indicates that the command scenario is in a critical stage of the mission, the iteration cycle is shortened based on the mission execution progress data. If the scenario is in a stable stage of the mission, the iteration cycle is extended.

10. An image data enhancement system for mobile command screens, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the image data enhancement method for a mobile command screen according to any one of claims 1 to 9 by executing the machine-executable instructions.

Images