Battlefield situation data label matching device and terminal equipment

Abstract

The invention relates to the technical field of real-time data processing, in particular to a battlefield situation data label matching device and terminal equipment, and the device comprises an own unit identification module, a position identification matching module, a battlefield event identification module, a real-time situation analysis module, a voice information analysis module and a label verification and updating module. According to the method, the battlefield visual data is captured and analyzed in real time, the type and state of a unit are recognized in real time, the timeliness of information and the continuity of dynamic tracking are ensured, automatic matching of geographic tags and real-time classification of event types are combined, the analysis ability of a complex battlefield environment is enhanced, and the analysis efficiency of the battlefield environment is improved. Resource configuration optimization and tactical deployment are realized, through automatic analysis of voice information and matching of action tags, command efficiency is improved, information delay and errors are reduced, a tag verification and updating mechanism is utilized, accuracy and consistency of data tags are ensured, and timeliness and integrity of data in a rapidly changing battlefield environment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of real-time data processing, and in particular to a matching device and terminal equipment for battlefield situation data labels. Background Art

[0002] The field of real-time data processing technology focuses on collecting, processing and analyzing data from multiple sources in real time. It supports applications that require rapid decision-making by quickly processing batch data streams and providing instant output and insights. It is particularly critical in multiple industries such as military, finance, medical care, and network security. It aims to monitor and respond to real-time changes in the operating environment, including market dynamics, network intrusion detection, and battlefield situation changes. It combines complex event processing and stream processing frameworks to achieve continuous query and instant analysis of data, ensure data accuracy and processing speed, match the requirements of multiple application scenarios, and enable decision makers to respond based on the latest and most comprehensive data information.

[0003] Among them, a matching device for battlefield situation data labels is used to improve data processing efficiency and accuracy in a military battlefield environment. It matches and verifies label information in battlefield data in an automated manner, including troop positions, equipment status, and enemy and friendly activities. It aims to ensure that the information provided to commanders is accurate and timely under rapidly changing battlefield conditions, enhance military commanders' immediate understanding of battlefield situations, optimize tactical decisions and resource allocation, and achieve effective strategic decisions and tactical deployments. By updating and verifying data in real time, it improves the response speed and operational efficiency of battlefield command, reduces human errors, improves data processing efficiency, and achieves higher operational accuracy and response speed in a battlefield environment.

[0004] Traditional battlefield situation data label matching technology has deficiencies in multi-source data fusion and instant updating. It is difficult to process heterogeneous data from multiple sensors and communication systems, which limits the information integration capability and processing speed. Without fully utilizing visual recognition and geographic labeling technology, it is impossible to identify and classify complex battlefield events in real time, resulting in tactical decisions based on data that is not accurate enough or not updated in a timely manner. As a result, commanders are unable to respond quickly to battlefield changes, missing tactical opportunities or making errors in resource deployment. Voice information processing usually cannot be matched with battlefield data in real time, limiting the improvement of command efficiency. In terms of label consistency and timeliness, the lack of an effective automatic verification mechanism often leads to decisions based on outdated or inaccurate data, increasing the risk and uncertainty of battlefield operations. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose a matching device and terminal equipment for battlefield situation data labels.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a matching device for battlefield situation data labels comprises: The friendly unit recognition module analyzes the visual features and behavior patterns of various units based on battlefield visual image data, identifies the type and real-time status of friendly units, and generates unit status recognition results; The location recognition and matching module uses the unit status recognition results to analyze the battlefield image, identify geographical features, and annotate geographical elements in combination with the location information to generate a geographical label data set; The battlefield event recognition module uses the geo-tagged data set to analyze battlefield images, recognize and classify various battlefield events, and generate event type classification results; The real-time scenario analysis module analyzes battlefield changes in real time, identifies battlefield scenarios, and generates scenario label matching results based on the event type classification results, combined with geographic location information, weather, terrain changes, and unit activity data; The voice information analysis module identifies the content of the voice call, matches the command and action tags, and generates a call voice analysis result based on the scenario tag matching result through voice recognition and semantic analysis; The label verification and update module checks the logical consistency and timeliness of multiple labels based on the call voice analysis results, identifies erroneous and outdated labels and updates them, and generates data label matching results.

[0007] As a further solution of the present invention, the unit status identification result includes unit type information, equipment status identification information, and activity timestamp; the geo-tag data set includes terrain contour identification results, building location data, and associated geographic coordinate information; the event type classification result includes explosion event identification information, shooting event identification record information, and unit movement path identification results; the scenario label matching result includes weather change labels, terrain change labels, and unit activity scenario labels; the call voice analysis result includes the identified command content, action label matching records, and voice command time information; the data label matching result includes unit label verification records, event label correction results, and geo-tag update results.

[0008] As a further solution of the present invention, the own unit identification module includes: The image feature extraction submodule analyzes and identifies the visual features of various units on the battlefield, including equipment types and iconic symbols, based on battlefield visual image data, and performs visual data analysis to obtain visual feature data; The behavior pattern recognition submodule evaluates and recognizes the behavior pattern of the unit based on the visual feature data by analyzing multiple movement speeds and activity routes in the battlefield, and generates a unit behavior analysis result; The real-time status recording submodule records the personnel and equipment status information of the own unit in real time based on the unit behavior analysis results, including the number of personnel and equipment operation status, and generates a unit status recognition result in combination with time information.

[0009] As a further solution of the present invention, the position identification and matching module includes: The terrain feature analysis submodule analyzes battlefield images in real time based on the unit status recognition results, calculates the total difference value of terrain features, recognizes multiple terrain features, including mountains, rivers, and valleys, adjusts the image resolution and contrast of the terrain to optimize the recognition accuracy, and generates terrain recognition results; The building structure marking submodule analyzes the battlefield image based on the terrain recognition result, identifies buildings and various man-made structures, including bunkers and military camps, marks the outline of the building and matches the label information, and generates a building recognition result; The geo-tag matching submodule, based on the building recognition result and in combination with the geographical location information, performs location annotation on the recognized geographical elements, including location coordinates and geographical feature information, and generates a geo-tag data set.

[0010] As a further solution of the present invention, the specific formula for calculating the total difference value of the terrain features is: ; Among them, D represents the total difference value of the terrain features, which is used to measure the significance of the terrain features in the entire image. Represents the LBP value of the kth local area extracted from the image, reflecting the terrain texture characteristics of the area. K is the total number of local areas considered in the image, indicating the number of local samples involved in terrain analysis. k is an index used to traverse all local areas. The formula classifies and identifies terrain features by calculating the sum of the absolute differences between the LBP value of each local area and the overall average LBP value. By comprehensively evaluating the texture information of each local area, various terrain types are identified and the accuracy of terrain recognition is improved.

[0011] As a further solution of the present invention, the battlefield event recognition module includes: The image information analysis submodule analyzes the battlefield image based on the geo-tagged data set, identifies event features in the image, including explosion shapes, shooting traces, and unit movements, and generates event feature analysis results; The event category analysis submodule classifies various types of battlefield event features based on the event feature analysis results, including explosion, shooting, unit movement and multiple categories, and generates feature type analysis data; The event label matching submodule analyzes the data based on the feature type, combines the time, location and unit type information of multiple events, matches labels for multiple identified and classified events, and generates event type classification results.

[0012] As a further solution of the present invention, the real-time scenario analysis module includes: The scenario data integration submodule generates a scenario data set based on the event type classification results, combined with geographic location information and real-time battlefield images, according to weather, terrain changes and unit activity data; The scenario change analysis submodule analyzes the dynamic changes of the battlefield in real time based on the scenario data set, including terrain changes, weather changes, unit assembly, and dispersed areas, and generates scenario analysis results; The scenario type classification submodule generates scenario label matching results by identifying and classifying battlefield scenarios based on the scenario analysis results, matching label information for battlefield scenarios in multiple regions.

[0013] As a further solution of the present invention, the voice information analysis module includes: The speech feature extraction submodule analyzes the speech signal based on the scenario label matching result and battlefield communication data, and performs spectrum analysis and amplitude adjustment on the signal to generate speech feature data; The semantic content analysis submodule identifies the command and instruction content in the voice signal through semantic analysis based on the voice feature data, extracts keywords and command information, and generates a semantic analysis result; The command matching submodule matches the recognized command with a preset action tag database based on the semantic analysis result to generate a call voice analysis result.

[0014] As a further solution of the present invention, the label verification and update module includes: The consistency check submodule performs a logical consistency check on the generated multiple tags based on the call voice analysis results, analyzes the consistency between the tags by comparing the tag data of multiple data sources and time points, and generates tag consistency evaluation information; The time information analysis submodule verifies the timeliness of the label data based on the label consistency evaluation information, identifies erroneous and outdated labels, records label items that need to be updated and corrected, and generates timeliness verification results; The label update submodule updates the label information based on the timeliness verification result, corrects errors and optimizes the label data in the database, optimizes the timeliness of the label information, and generates a data label matching result.

[0015] A battlefield situation data tag matching terminal device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the content of the battlefield situation data tag matching device when executing the computer program.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by real-time capture and analysis of battlefield visual data, real-time identification of unit types and states is achieved, ensuring the timeliness of information and the continuity of dynamic tracking. Combined with automatic matching of geographic tags and real-time classification of event types, the ability to analyze complex battlefield environments is enhanced, resource allocation optimization and tactical deployment are achieved, and command efficiency is improved and information delays and errors are reduced through automatic parsing of voice information and matching of action tags. The accuracy and consistency of data tags are ensured by using a tag verification and update mechanism, thereby improving the timeliness and integrity of data in a rapidly changing battlefield environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of the device of the present invention; Figure 2 It is a schematic diagram of the device framework of the present invention; Figure 3 This is a flow chart of the own unit identification module of the present invention; Figure 4 It is a flow chart of the location identification and matching module of the present invention; Figure 5 It is a flow chart of battlefield event identification module of the present invention; Figure 6 It is a flow chart of the real-time scenario analysis module of the present invention; Figure 7 It is a flow chart of the voice information analysis module of the present invention; Figure 8 This is a flow chart of the label verification and update module of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0019] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0020] See also Figure 1 to Figure 2 The present invention provides a technical solution, a matching device for battlefield situation data labels comprises: The friendly unit recognition module analyzes the visual features and behavior patterns of various units based on battlefield visual image data, identifies the type and real-time status of friendly units, and generates unit status recognition results; The location recognition and matching module uses the unit status recognition results to analyze the battlefield image, identify geographical features, and combine the location information to annotate the geographical elements to generate a geographical label data set; The battlefield event recognition module uses the geo-tagged dataset to analyze battlefield images, identify and classify a variety of battlefield events, and generate event type classification results; The real-time scenario analysis module analyzes battlefield changes in real time, identifies battlefield scenarios, and generates scenario label matching results based on event type classification results, combined with geographic location information, weather, terrain changes, and unit activity data; The voice information analysis module identifies the content of the voice call, matches the command and action tags, and generates the call voice analysis results based on the scenario tag matching results through voice recognition and semantic analysis; The label verification and update module checks the logical consistency and timeliness of multiple labels based on the call voice analysis results, identifies erroneous and outdated labels and updates them, and generates data label matching results.

[0021] The unit status identification results include unit type information, equipment status identification information, and activity timestamps. The geographic tag data set includes terrain contour identification results, building location data, and associated geographic coordinate information. The event type classification results include explosion event identification information, shooting event identification record information, and unit movement path identification results. The scenario label matching results include weather change labels, terrain change labels, and unit activity scenario labels. The call voice analysis results include the identified command content, action label matching records, and voice command time information. The data label matching results include unit label verification records, event label correction results, and geographic label update results.

[0022] See also Figure 2 and Figure 3 , the friendly unit identification module includes: The image feature extraction submodule analyzes and identifies the visual features of various units on the battlefield, including equipment types and iconic symbols, based on battlefield visual image data, and performs visual data analysis to obtain visual feature data; In the above content, the image feature extraction submodule collects battlefield visual image data and performs feature extraction of the deep convolutional neural network model according to the formula , calculate and obtain visual feature data; In the formula, F(v) represents the extracted visual feature result, represents the weight of the kth convolution kernel, Represents the response value of the input image data at the kth convolution kernel, b is the bias term, and K is the total number of convolution kernels; Detailed explanation of the formula and the process of formula calculation and derivation: Assume that a 3x3 convolution kernel is used for feature extraction, the convolution kernel weight w is assumed to be [0.2, -0.1, 0.05, 0.2, -0.2, 0.05, 0.1, 0.1, -0.15], and the bias b is set to 0.05. For the input image area A small area of ​​9 pixels has a response value of [100, 150, 100, 90, 130, 85, 120, 140, 95]. Calculate the feature extraction result of this area: ; The results show that in this image region, based on the weighted response and bias of the convolution kernel, the resulting feature value is 18.05, which will be used in subsequent classification or recognition tasks as part of the visual features of unit types and iconic symbols.

[0023] The behavior pattern recognition submodule evaluates and identifies the behavior patterns of units based on visual feature data by analyzing multiple movement speeds and activity routes in the battlefield, and generates unit behavior analysis results; In the behavior pattern recognition submodule, based on visual feature data, the movement speed and activity routes of multiple units on the battlefield are collected. Position tracking sensors and time series analysis technology are used to record the position changes of units within the target time. The behavior patterns of the units are identified by analyzing the data, including marching, stationary, and retreating. Algorithm processing includes speed calculation and path optimization. Abnormal behavior patterns are identified by comparing historical behavior data models. The patterns are evaluated through set behavior classification criteria to generate unit behavior analysis results. The results help military analysts understand the strategies and actions taken by both sides in the actual battlefield environment and generate unit behavior analysis results.

[0024] The real-time status recording submodule records the personnel and equipment status information of the own unit in real time based on the unit behavior analysis results, including the number of personnel and equipment operation status, and generates unit status recognition results in combination with time information; In the real-time status recording submodule, based on the unit behavior analysis results, the personnel and equipment status of one's own units are tracked and recorded in real time, including extracting key information from the behavior analysis results, such as the number of units and equipment type, combining real-time timestamp information, formatting and storing the data, and using the database management system to record and update status information. The records include the number of personnel, equipment type, and operating status of each unit, involving multiple states such as normal operation, maintenance required, or damaged. The information update is reflected on the monitoring interface for the command center to view and make decisions in real time. The generated status data provides military commanders with an instant understanding of the current battlefield situation and generates unit status identification results.

[0025] See also Figure 2 and Figure 4 , the location recognition and matching module includes: The terrain feature analysis submodule analyzes battlefield images in real time based on the unit status recognition results, calculates the total difference value of terrain features, recognizes various terrain features, including mountains, rivers, and valleys, adjusts the image resolution and contrast of the terrain to optimize the recognition accuracy, and generates terrain recognition results; The specific formula for calculating the total difference value of terrain features is: ; Among them, D represents the total difference value of the terrain features, which is used to measure the significance of the terrain features in the entire image. Represents the LBP value of the kth local area extracted from the image, reflecting the terrain texture characteristics of the area. K is the total number of local areas considered in the image, indicating the number of local samples involved in terrain analysis. k is an index used to traverse all local areas. The formula classifies and identifies terrain features by calculating the sum of the absolute differences between the LBP value of each local area and the overall average LBP value. By comprehensively evaluating the texture information of each local area, various terrain types are identified and the accuracy of terrain recognition is improved.

[0026] formula: ; Detailed explanation of the formula and the process of formula calculation and derivation: The formula is used to calculate the total difference value of terrain features. The result is used to evaluate the significance of terrain features in battlefield images. The calculation process includes summing the absolute difference between the LBP value of each local area and the overall average LBP value. Parameter meaning and setting value: K is the total number of local regions considered, which is assumed to be 5, indicating that the image is segmented into 5 main regions for independent analysis; is the local binary pattern value of the kth local area, which reflects the texture information of the terrain in each area. Assume that the value is , , , , ; is the average value of LBP values ​​in all local areas; ; Substitute the parameters into the formula for calculation: ; The result D=22 indicates the total difference value of terrain features, indicating that in these five areas, the texture differences of terrain features are more obvious. The larger the value, the more significant the difference between terrain features. The formula is used to ensure the accuracy of terrain recognition in complex battlefield environments and provide reliable data support for tactical decision-making.

[0027] The building structure labeling submodule analyzes battlefield images based on terrain recognition results, identifies buildings and various man-made structures, including bunkers and military camps, marks building outlines and matches label information to generate building recognition results; In the building structure marking submodule, based on the terrain recognition results, the battlefield images are analyzed using image processing algorithms to identify various types of buildings and man-made structures, including bunkers and military camps. The outlines of the buildings are determined through edge detection technology, and the image segmentation method is used to refine the visual boundaries of the building structures. The machine learning model convolutional neural network is used to identify and classify various building structures. The model training uses batch image data of battlefield buildings to improve recognition accuracy. The identified buildings are matched with corresponding label information through a preset label library, including type and purpose, to generate building recognition results including building name and category.

[0028] The geo-tag matching submodule, based on the building recognition results and combined with the geographic location information, annotates the identified geographic elements, including location coordinates and geographic feature information, and generates a geo-tag dataset; In the geo-tag matching submodule, based on the building recognition results and combined with the real-time geographic location information, the location of each identified building and geographic element is annotated, the type and location of the building are extracted from the building recognition results, and the information is merged with the map dataset using geographic information system technology. The relative position of the building is converted into geographic coordinates through a coordinate conversion algorithm, and additional geographic feature information is added to each geographic element, including terrain and surrounding environment. Data processing and spatial analysis techniques are used to ensure the accuracy of annotation and the integrity of geographic data, and a geo-tag dataset including complete location coordinates and geographic features is generated.

[0029] See also Figure 2 and Figure 5 ,The battlefield event recognition module includes: The image information analysis submodule analyzes battlefield images based on geo-tagged datasets, identifies event features in the images, including explosion shapes, shooting traces, and unit movements, and generates event feature analysis results; In the above content, the image information analysis submodule collects battlefield image data and executes feature detection algorithm according to the formula , calculate and obtain the event feature analysis results; In the formula, F(e) represents the extracted event feature result, represents the weight of the jth detection filter, represents the response value of the input image data in the jth filter, c is the bias term, and J is the total number of filters; Detailed explanation of the formula and the process of formula calculation and derivation: Assume that the filter weight w is set to [0.3, -0.2, 0.1], the bias c is set to 0.1, and for a 3-pixel feature in the input image region e, the response value is [200, 150, 180]. Calculate the event feature extraction result of this region: ; The results show that in this image region, the eigenvalue obtained based on the weighted response and bias of the feature detection filter is 48.1, and the value is used in the event recognition task as part of the explosion shape, shooting traces, and unit movement.

[0030] The event category analysis submodule classifies various types of battlefield event features based on the event feature analysis results, including explosions, shooting, unit movement, and other categories, and generates feature type analysis data; In the event type classification submodule, based on the event feature analysis results, a classification algorithm is used to distinguish the types of various events occurring on the battlefield, including explosions, shootings, unit movements and other categories. Each event feature data will be loaded into the classification model. The model can recognize the characteristic patterns of different events based on the trained data set. Explosion events are identified based on their shape and diffusion pattern, shooting events are classified according to shooting traces and directions, and unit movement events are identified through dynamic changes and path predictions. Each event type is classified through a corresponding standardization process to ensure consistency and accuracy. Through multiple processing steps, different types of battlefield events are classified and feature type analysis data is generated.

[0031] The event label matching submodule analyzes data based on feature types, combines the time, location, and unit type information of multiple events, matches labels for multiple identified and classified events, and generates event type classification results; In the event label matching submodule, based on the feature type analysis data, combined with the time, location and type information of the participating units, appropriate labels are matched for different battlefield events. Through the comprehensive use of database management and information retrieval technology, key event data, including event type, time and location of occurrence, are extracted from the classification results, and compared and matched with the pre-set event label library. The specific circumstances and background of the event are taken into account during the label matching process to ensure the accuracy and relevance of each label. Through data comparison and label logic judgment, one or more labels are assigned to each event to reflect the characteristics and impact of the event and generate event type classification results.

[0032] See also Figure 2 and Figure 6 , the real-time scenario analysis module includes: The scenario data integration submodule generates scenario data sets based on the event type classification results, combined with geographic location information and real-time battlefield images, according to weather, terrain changes and unit activity data; In the above content, the scenario data integration submodule integrates the event type classification results and geographic location information, applies the data fusion algorithm, and uses the formula , calculate and generate scenario data sets; Where S represents the score of the integrated scenario dataset, which is used to measure the comprehensive importance of geographic and battlefield image features. and are the weight coefficients of geographical location and battlefield image features, and are the corresponding geographic and image data points, respectively. is the bias term used to adjust the integration results; Detailed explanation of the formula and calculation process: Assume that the geographic location data point is set to =10, =20, =15, the battlefield image data points are set as , , , the weight coefficient a is , , , the weight coefficient b is , , , constant term is 5, calculate the scenario dataset score for this area: ; The result S = 52.5 shows that based on the comprehensive evaluation of weights and data, the obtained scenario importance score is 52.5. The score reflects the relative importance and urgency of the features extracted from the fusion of geographic and battlefield image information, which is used for battlefield scenario analysis and guiding the decision-making process. The response strategy to different battlefield information is optimized through weight adjustment to enhance battlefield resilience.

[0033] The scenario change analysis submodule analyzes the dynamic changes of the battlefield in real time based on the scenario data set, including terrain changes, weather changes, unit assembly, and dispersed areas, and generates scenario analysis results; In the scenario change analysis submodule, based on the scenario data set, the dynamic changes of the battlefield are monitored and analyzed in real time, including changes in terrain, changes in weather conditions, and the assembly and dispersion of units. Dynamic models and real-time data stream analysis technology are used to continuously track battlefield changes. The model analyzes the time tags and geographic tags in the data set to identify key change points, including specific areas where the terrain is destroyed or drastic changes in weather conditions. The assembly and dispersion of units are also analyzed through location data and movement patterns to determine potential strategic adjustments and response measures, helping military decision makers understand the current battlefield operating environment and generate scenario analysis results to describe dynamic changes on the battlefield.

[0034] The scenario type classification submodule generates scenario label matching results by identifying and classifying battlefield scenarios based on the scenario analysis results, matching label information for battlefield scenarios in multiple regions; In the scenario type classification submodule, based on the scenario analysis results, different scenarios on the battlefield are identified and classified by using a classification algorithm, and scenario labels are matched for multiple areas on the battlefield. Each scenario is preliminarily classified according to its geographical location, involved units and event characteristics. The machine learning models support vector machine and random forest are used to classify the scenario types. The classification is based on a preset labeling system, including a variety of scenario labels from a strategic perspective, including "high risk", "active conflict zone" or "reconstruction zone". By considering the complexity of the scenario and multi-dimensional data input, the accurate matching of labels is ensured, and the classification results of scenario types are generated to provide accurate regional battlefield scenario descriptions for military command.

[0035] See also Figure 2 and Figure 7 , the voice information analysis module includes: The speech feature extraction submodule analyzes the speech signal based on the scenario label matching results and battlefield communication data, and performs spectrum analysis and amplitude adjustment on the signal to generate speech feature data; In the speech feature extraction submodule, based on the scenario label matching results, battlefield communication data is used to analyze speech signals. The signals are preprocessed, including noise suppression and signal enhancement, to improve speech quality. Next, spectrum analysis is performed, and the time domain signal is converted to the frequency domain using fast Fourier transform. The frequency components of the signal are analyzed, and the signal amplitude is adjusted to ensure the clarity of speech in different environments. Clear speech features, including pitch, intensity, and rhythm, are extracted from the complex battlefield environment. The feature data is encoded and stored to provide the necessary input for semantic analysis and generate speech feature data.

[0036] The semantic content analysis submodule is based on the voice feature data, identifies the commands and instructions in the voice signal through semantic analysis, extracts keywords and command information, and generates semantic analysis results; In the semantic content analysis submodule, natural language processing technology is used to perform semantic analysis on voice signals based on voice feature data, and voice data is transcribed into text through voice recognition technology. Text analysis algorithms, including named entity recognition and keyword extraction, are used to determine the key content in the voice, including commands and instructions. The content is processed through semantic understanding to identify action instructions and battlefield-related terms, and generate semantic parsing results including instructions and key decision words to support subsequent command matching and action decisions.

[0037] The command matching submodule matches the recognized command with the preset action tag database based on the semantic analysis results to generate the call voice analysis results; In the command matching submodule, the identified command content is matched based on the semantic parsing results, and the action labels associated with the semantic parsing results are retrieved from the preset action label database. The database includes a variety of commands and corresponding action instructions. The command text and the action labels in the database are compared through the matching algorithm to ensure the correct interpretation and application of the command, and the voice instructions on the battlefield are converted into action plans. The generated call voice analysis results reflect the content of the voice commands and the predetermined action responses, providing accurate execution guidelines for military command.

[0038] See also Figure 2 and Figure 8 , the label verification and update module includes: The consistency check submodule performs logical consistency checks on the various labels generated based on the call voice analysis results. By comparing label data from multiple data sources and time points, it analyzes the consistency between labels and generates label consistency evaluation information. In the above content, the logical consistency of the generated multiple tags is checked according to the formula , calculate label consistency assessment information; In the formula, C represents the consistency score, n represents the total number of compared labels, and Represent the same label item in the new and old data sources respectively; Detailed explanation of the formula and the process of formula calculation and derivation: Assume there are three label items, the label values ​​of the new data source are [100, 200, 300], and the label values ​​of the old data source are [95, 195, 305]. Calculate the consistency score: ; The results show that the average consistency deviation between labels is 5, which means that the new and old label data have high consistency within the considered error range, and the need for updating or changing the label system is low.

[0039] The time information analysis submodule verifies the timeliness of label data based on label consistency assessment information, identifies incorrect and outdated labels, records label items that need to be updated and corrected, and generates timeliness verification results; In the time information analysis submodule, the label data is verified for timeliness based on the label consistency evaluation information, including checking the creation and last update time of each label, identifying erroneous or outdated labels through the time analysis algorithm, and determining whether the current status of the label reflects the latest information by analyzing the label's validity period and data update frequency, identifying label items that need to be updated or corrected, and verifying the timeliness status of multiple labels to evaluate whether they need to be updated or remain unchanged, thereby obtaining a timeliness verification result.

[0040] The label update submodule updates label information based on the timeliness verification results, corrects errors and optimizes label data in the database, optimizes the timeliness of label information, and generates data label matching results; In the label update submodule, based on the timeliness verification results, the label information is updated and optimized, and the labels identified as erroneous or outdated are analyzed to determine the causes of the errors and obsolete links. According to the analysis results, the errors are corrected by updating the algorithm and database operations, and the label data is optimized to ensure that the information of each label is up to date and accurate. Database management and data processing technologies are used to correct errors, improve the label structure and storage efficiency within the database, generate data label matching results, and ensure that the label information in the database is accurate and timely.

[0041] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A battlefield situation data label matching device, characterized in that: The device comprises: The friendly unit recognition module analyzes the visual features and behavior patterns of various units based on battlefield visual image data, identifies the type and real-time status of friendly units, and generates unit status recognition results; The location recognition and matching module uses the unit status recognition results to analyze the battlefield image, identify geographical features, and annotate geographical elements in combination with the location information to generate a geographical label data set; The battlefield event recognition module uses the geo-tagged data set to analyze battlefield images, recognize and classify various battlefield events, and generate event type classification results; The real-time scenario analysis module analyzes battlefield changes in real time, identifies battlefield scenarios, and generates scenario label matching results based on the event type classification results, combined with geographic location information, weather, terrain changes, and unit activity data; The voice information analysis module identifies the content of the voice call, matches the command and action tags, and generates a call voice analysis result based on the scenario tag matching result through voice recognition and semantic analysis; The label verification and update module checks the logical consistency and timeliness of multiple labels based on the call voice analysis results, identifies erroneous and outdated labels and updates them, and generates data label matching results.

2. The battlefield situation data tag matching device according to claim 1, characterized in that: The unit status identification result includes unit type information, equipment status identification information, and activity timestamp; the geo-tag data set includes terrain contour identification results, building location data, and associated geographic coordinate information; the event type classification result includes explosion event identification information, shooting event identification record information, and unit movement path identification result; the scenario label matching result includes weather change labels, terrain change labels, and unit activity scenario labels; the call voice analysis result includes the identified command content, action label matching records, and voice command time information; the data label matching result includes unit label verification records, event label correction results, and geo-tag update results.

3. The battlefield situation data tag matching device according to claim 1, characterized in that: The own unit identification module includes: The image feature extraction submodule analyzes and identifies the visual features of various units on the battlefield, including equipment types and iconic symbols, based on battlefield visual image data, and performs visual data analysis to obtain visual feature data; The behavior pattern recognition submodule evaluates and recognizes the behavior pattern of the unit based on the visual feature data by analyzing multiple movement speeds and activity routes in the battlefield, and generates a unit behavior analysis result; The real-time status recording submodule records the personnel and equipment status information of the own unit in real time based on the unit behavior analysis results, including the number of personnel and equipment operation status, and generates a unit status recognition result in combination with time information.

4. The battlefield situation data tag matching device according to claim 1, characterized in that: The location identification and matching module comprises: The terrain feature analysis submodule analyzes battlefield images in real time based on the unit status recognition results, calculates the total difference value of terrain features, recognizes multiple terrain features, including mountains, rivers, and valleys, adjusts the image resolution and contrast of the terrain to optimize the recognition accuracy, and generates terrain recognition results; The building structure marking submodule analyzes the battlefield image based on the terrain recognition result, identifies buildings and various man-made structures, including bunkers and military camps, marks the building outlines and matches the label information, and generates a building recognition result; The geo-tag matching submodule, based on the building recognition result and in combination with the geographical location information, performs location annotation on the recognized geographical elements, including location coordinates and geographical feature information, and generates a geo-tag data set.

5. The battlefield situation data tag matching device according to claim 4, characterized in that: The specific formula for calculating the total difference value of terrain features is: ； Among them, D represents the total difference value of the terrain features, which is used to measure the significance of the terrain features in the entire image. Represents the LBP value of the kth local area extracted from the image, reflecting the terrain texture characteristics of the area. K is the total number of local areas considered in the image, indicating the number of local samples involved in terrain analysis. k is an index used to traverse all local areas. The formula classifies and identifies terrain features by calculating the sum of the absolute differences between the LBP value of each local area and the overall average LBP value. By comprehensively evaluating the texture information of each local area, various terrain types are identified and the accuracy of terrain recognition is improved.

6. The battlefield situation data tag matching device according to claim 1, characterized in that: The battlefield event recognition module includes: The image information analysis submodule analyzes the battlefield image based on the geo-tagged data set, identifies event features in the image, including explosion shapes, shooting traces, and unit movements, and generates event feature analysis results; The event category analysis submodule classifies various types of battlefield event features based on the event feature analysis results, including explosion, shooting, unit movement and multiple categories, and generates feature type analysis data; The event label matching submodule analyzes the data based on the feature type, combines the time, location and unit type information of multiple events, matches labels for multiple identified and classified events, and generates event type classification results.

7. The battlefield situation data tag matching device according to claim 1, characterized in that: The real-time scenario analysis module includes: The scenario data integration submodule generates a scenario data set based on the event type classification results, combined with geographic location information and real-time battlefield images, according to weather, terrain changes and unit activity data; The scenario change analysis submodule analyzes the dynamic changes of the battlefield in real time based on the scenario data set, including terrain changes, weather changes, unit assembly, and dispersed areas, and generates scenario analysis results; The scenario type classification submodule identifies and classifies battlefield scenarios based on the scenario analysis results, matches label information for battlefield scenarios in multiple regions, and generates scenario label matching results.

8. The battlefield situation data tag matching device according to claim 1, characterized in that: The voice information analysis module comprises: The speech feature extraction submodule analyzes the speech signal based on the scenario label matching result and battlefield communication data, and performs spectrum analysis and amplitude adjustment on the signal to generate speech feature data; The semantic content analysis submodule identifies the command and instruction content in the voice signal through semantic analysis based on the voice feature data, extracts keywords and command information, and generates a semantic analysis result; The command matching submodule matches the recognized command with a preset action tag database based on the semantic analysis result to generate a call voice analysis result.

9. The battlefield situation data tag matching device according to claim 1, characterized in that: The label verification and update module includes: The consistency check submodule performs a logical consistency check on the generated multiple tags based on the call voice analysis results, analyzes the consistency between the tags by comparing the tag data of multiple data sources and time points, and generates tag consistency evaluation information; The time information analysis submodule verifies the timeliness of the label data based on the label consistency evaluation information, identifies erroneous and outdated labels, records label items that need to be updated and corrected, and generates timeliness verification results; The label update submodule updates the label information based on the timeliness verification result, corrects errors and optimizes the label data in the database, optimizes the timeliness of the label information, and generates a data label matching result.

10. A battlefield situation data tag matching terminal device, comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor implements the battlefield situation data label matching device described in any one of claims 1 to 9 when executing the computer program.

Images