136 results about "Scene analysis" patented technology

The invention discloses a big data platform asset intelligent sensing method based on LLM and a customizable MCP, and the method comprises the steps: defining a standard resource library of a model context protocol, constructing a model context protocol server, integrating a plurality of scanning tools, and carrying out the automatic adaption; receiving a natural language demand input by a user, and calling a pre-trained LLM model; the scene analysis module converts a natural language demand into a standardized scene description object; the intelligent scanning module is used for describing an object according to a standardized scene, generating an optimal tool execution chain, executing a complete scanning task, obtaining a task scanning result and unifying formats of heterogeneous data in the task scanning result, and the conflict resolution module is used for removing conflicts; the multi-source data association module fuses network topology resources and flow data captured by a probe to construct a dynamic asset atlas including asset attributes, service dependence and vulnerability information, and the report generation module generates a structured report. The method can provide a global risk perspective and decision support.

The invention relates to the field of image processing, and discloses an object monitoring method and device based on multi-source data and a storage medium, and the method comprises the steps: synchronously collecting multi-channel video streams, environment parameters and equipment position information, and carrying out the time alignment and data association processing, and forming associated data; performing feature matching on the multiple paths of video streams, fusing position information and environment parameters, and establishing a mapping relation between video feature points and a unified space coordinate system; splicing the multiple paths of video streams in real time according to the mapping relation to generate a panoramic video stream; performing dynamic scene analysis based on the panoramic video stream and the environmental parameters, and identifying a target type and a state to obtain an analysis result; and generating an equipment regulation and control strategy based on the analysis result, and generating and issuing a regulation and control instruction for adjusting the working parameters of the multi-view image acquisition device based on the equipment regulation and control strategy. According to the invention, the splicing precision, the real-time performance and the target identification accuracy of panoramic monitoring can be improved, and the dynamic adaptive regulation and control of the equipment can be realized.

The invention particularly relates to an atomic business intelligent arrangement device and method. The atomic business intelligent arrangement device comprises an atomic business relation graph intelligent construction module, an atomic business fine management module, a business intelligent arrangement and visual arrangement module and a man-machine interaction module. According to the atomic service intelligent arrangement device and method, full-chain intelligentization from user demand identification, service scene analysis to process automatic arrangement is realized, the service handling efficiency and the service quality are improved, manual correction is supported, the error rate is reduced, the reusability of atomic services is improved, policy scenes can be helped to rapidly fall into the ground, and the service quality is improved. And diversified operation requirements of the user are met, so that the user satisfaction is improved.

The invention provides an environmental protection planning dynamic generation method and system based on a multi-modal large model. The method comprises the following steps: firstly, obtaining environmental protection planning target scene demand information covering text and image forms; calling the multi-modal large model to carry out cross-modal scene analysis, and generating an analysis result containing the core demand elements and priority ranking; generating a preliminary environmental protection planning scheme through a cross-modal planning content generation module based on the analysis result; collecting scheme pilot run feedback information and inputting the information into the multi-modal model for feedback adaptation analysis, and generating an adjustment rule; and finally, dynamically adjusting the initial scheme according to an adjustment rule to obtain a final scheme. According to the invention, scene requirements can be comprehensively and accurately captured, a scientific planning scheme is accurately generated, dynamic optimization is realized, and environmental protection planning efficiency and quality are improved.

The invention discloses a new-generation emergency command system scene analysis system integrated with a convolutional neural network, particularly relates to the technical field of computer vision and intelligent emergency command, and is used for solving the problem that the dynamic association and situation evolution logic analysis capability among emergency scene elements is insufficient in the prior art. Visual element features are extracted from a video stream through a feature extraction module by using a convolutional neural network, an association graph construction module constructs a space-time association graph based on the features, a movement evaluation module calculates element group movement feasibility indexes, and a situation analysis module analyzes divergence features and antagonism degrees in combination with movement track convergence. And the risk identification module identifies a derived risk aggregation mode according to the situation prediction, and the semantic generation module finally outputs high-level semantic description information containing situation prediction to realize closed-loop analysis from scene perception to cognitive decision.

The invention belongs to the technical field of data processing, and particularly discloses a computer vision model service scenarized configuration generation method, which comprises the following steps: constructing a basic information database comprising a plurality of computer vision models, the database being used for storing model identification information and feature parameters; extracting structured features of the scene according to a scene analysis engine, wherein the structured features are used for describing operation conditions of the scene; calculating a model matching score according to the structured features and the feature parameters, and obtaining a recommendation model through a multi-condition weight matching function; executing a model selection strategy based on the matching score, and obtaining a model matched with the scene; and generating a configuration file conforming to a predetermined specification according to the recommendation model. The objective of the invention is to solve the problems of low efficiency and error proneness caused by the fact that a computer vision model in the prior art depends on a manual selection model and cannot quickly select and adapt to an optimal model in diversified scenes.

The invention discloses an enhanced GAN extreme scene generation method and system considering weather factors, and relates to the technical field of scene generation in new energy power system planning and operation, and the method comprises the steps: extracting key factors from multi-time-delay meteorological data, forming an extreme label, converting the extreme label into a generation condition, and achieving the output through two-stage confrontation generation, wherein a scarce sample is stably amplified by WGAN-GP, a joint curve is generated according to conditions, and related structures, climbing and power boundary constraints are synchronously embedded to ensure statistical and physical consistency. According to the method, the authenticity of the distribution tail, the cross-variable correlation and the time structure can be considered when the sample is scarce. Extreme situation deduction of a novel power system can be supported while extreme situation coverage and credibility are improved, and an existing new energy scene analysis and planning process is integrated.

A container simulation network system based on a link autonomous agent is composed of a visual interaction module, a control plane module, a network simulation module and a resource management module. The modules perform information interaction and task collaboration through a uniform interface so as to realize automatic construction and management of the simulation network, a two-layer architecture design of network bridge containers and user containers is adopted, and each user container corresponds to one network bridge container; a link autonomous agent module is deployed in the network bridge container and is used for collecting a link state, calculating a performance adaptation degree and executing a scheduling operation bit; the visual interaction module realizes visual monitoring and interaction control of a simulation state; the control plane module is responsible for simulation scene analysis, link parameter generation and global scheduling management; and the network simulation module is responsible for calling a container generation tool to generate a specific container instance and carrying out resource monitoring and adjustment on the container.

The invention discloses an immersion type 3D scene intelligent optimization generation system and method based on AIGC, and particularly relates to the crossing field of computer graphics and artificial intelligence technologies. The system comprises a scene analysis and semantic modeling module, a style control module, a semantic interactive editing module, an immersive feedback acquisition module and an intelligent optimization loop module. According to the core method, a scene semantic map is constructed by analyzing user input, and scene visual consistency is guaranteed by using a global style control vector; realizing incremental updating driven by a local instruction through a semantic graph interface; and scene automatic iterative optimization is driven by fusing user behavior data and dominant feedback. According to the method and the system, the full-process intelligentization from generation, editing to optimization is realized, and the efficiency, the consistency and the user intention fitting degree of 3D scene creation are improved.

The invention relates to a medical quality control dynamic scoring system, device and equipment based on multi-modal perception and federal evolution, and the method comprises the steps: obtaining original scene information through multi-modal input based on a medical equipment recognition model, carrying out the scene analysis and semantic fusion, generating structured medical scene data, and carrying out the dynamic scoring of the medical quality control according to the structured medical scene data. Multi-source evidence verification is carried out through a medical scoring rule base, a quality evaluation result is generated through weight dynamic adjustment, difference analysis is carried out through comparison with a manual correction result according to the quality evaluation result, and model parameters and scoring rule updating are generated. Multi-modal input is fused based on a medical special perception layer, anti-interference scene analysis and equipment semantic binding are achieved, multi-source evidences are verified in real time through a dynamic scoring engine, deduction weights are dynamically adjusted, subjective item scoring floating is compressed, an identification model is updated through federation, and a rule base is thermally loaded in real time, so that the anti-interference scene analysis and equipment semantic binding is achieved. And automatic closed-loop processing of problem items is realized.

The invention provides a compliance strategy generation and risk early warning method and system based on multi-agent cooperation, and relates to the field of artificial intelligence and enterprise compliance management, and the method comprises the steps: collecting compliance basis documents through an information arrangement agent, and constructing a structured knowledge base; monitoring the change of the compliance basis document in the knowledge base through an information updating agent; receiving a business scene described by a user in a natural language through a scene analysis agent, and retrieving a compliance basis related to the business scene from a knowledge base; converting the retrieved compliance basis into an executable logic rule through a strategy generation agent; and acquiring real-time operation data from the business system through the risk early warning agent, comparing the operation data with the logic rule, and triggering risk early warning if violation occurs. According to the invention, automatic extraction, structured arrangement, dynamic updating, intelligent association and real-time judgment of compliance knowledge are realized, and the intelligent level and operation efficiency of enterprise compliance management are improved.

The invention relates to the technical field of AI interaction, in particular to a user portrait construction and personalized interaction system based on an AI dynamic prompt project, which comprises a scene analysis module used for recognizing an interaction scene and a core optimization target in the scene according to user input; the co-processing module is used for carrying out quantitative evaluation and sorting on the long-term features and the real-time features according to respective stability parameters of the long-term features and the real-time features and correlation parameters of the long-term features and the real-time features and the core optimization target, and generating a hierarchical feature set; the instruction generation module is used for adapting an instruction template according to the interaction scene and fusing the hierarchical feature set into the instruction template to generate a structured driving instruction; and the interaction execution module is used for executing the driving instruction to generate and output final personalized interaction content. According to the method, accurate balance of long and short term features of the user in personalized prompt generation can be realized through a scenarized dynamic weight distribution and feature conflict coordination mechanism.

This invention discloses a low-light scene analysis method based on multimodal feature fusion and clustering, belonging to artificial intelligence technology. It constructs a single-branch feature extraction network based on Transformer; a multimodal feature interaction and fusion module to achieve multimodal feature interaction and fusion; a multimodal fusion feature clustering module to cluster the features after multimodal interaction and fusion, utilizing the semantic and spatial distances between different features to achieve clustering while simultaneously downsampling the features, mitigating the problem of loss of detailed edge information in structured downsampling; and a multi-scale feature aggregation and decoding module to receive feature information from the encoding network and classify each feature pixel according to the semantic distance of the multi-scale features. This invention can fully utilize visible light and thermal image information and can be applied to scene analysis and navigation of unmanned systems in diverse and complex low-light scenes.

The high-speed train defect identification method and device, equipment and medium provided by the application relate to the technical field of computer vision. In the application, firstly, global scene analysis is performed on a target high-speed train image to obtain image scene description data; secondly, based on the image scene description data, local anomaly analysis is performed on the target high-speed train image to obtain a candidate abnormal region set; then, anomaly semantic verification is performed on each candidate abnormal region in the candidate abnormal region set to obtain a defect identification result of each candidate abnormal region; finally, based on the defect identification result of each candidate abnormal region, a target abnormal region is determined in the candidate abnormal region set, and spatial positioning processing is performed on the target abnormal region to obtain a spatial positioning result. Based on the above, the problem that the reliability of defect identification is relatively low in the prior art can be improved.

The application discloses a kind of earphone anti-misoperation control method and system based on fingerprint identification, it is related to the field of fingerprint identification, including, user static fingerprint data and dynamic fingerprint data are collected and preprocessed and stored in fingerprint database;Based on the real-time acquisition user dynamic fingerprint data, generate elliptical dynamic fingerprint window;Elliptical dynamic fingerprint window is converted into rectangular dynamic fingerprint window, and fingerprint feature point matching is carried out;User earphone use data is collected, and touch scene analysis model is constructed, and earphone use scene is obtained;According to earphone use scene, dynamically adjust the operation logic of earphone.The application generates rectangular dynamic fingerprint window, accurately captures the dynamic change of user finger in the touch process, can adapt to the change of finger in different operation, realizes the effective tracking and identification of real-time dynamic fingerprint, improves the accuracy of fingerprint matching, reduces misoperation, and enhances the adaptive capacity of system to complex touch operation.

The invention relates to the technical field of knowledge maps, can be applied to an intelligent traffic operation management scene, and discloses an intelligent decision-making method, device, equipment and medium based on a semantic knowledge map, and the method comprises the steps: carrying out the analysis, cleaning and complementation of multi-source heterogeneous data, and generating a standardized event flow under a unified space-time reference; constructing a semantic knowledge graph containing entities, attributes, relationships and business logic, injecting the standardized event streams into the graph to complete pattern matching, generating a structured knowledge packet and realizing business logic updating; and scene analysis is carried out based on the structured knowledge packet and the semantic knowledge graph, an analysis result is output, a decision suggestion is generated, and closed-loop processing from data to decision is realized. According to the method, multi-source data is converted into structured knowledge and semantic reasoning is introduced, so that the analysis result can support decision generation, and the semantic understanding ability and decision support efficiency of data fusion are improved.

The application provides an automatic driving rear-end collision and sudden state prediction method based on space-time occupation characteristics, and relates to the field of automatic driving.The application proposes a prediction network based on a space-time graph Transformer, solves the problem of lack of space-time occupation characteristics in automatic driving vehicle dangerous scene analysis.Through integration of multi-dimensional characteristics from a nuScenes data set, a rear-end near-collision scene knowledge graph RNSKG is constructed, and DBSCAN unsupervised clustering is used to classify dangerous states and emergency states.A space-time graph Transformer model coupled with a time sequence and a space attention mechanism is significantly better than a traditional model in predicting dangerous states and emergency states, verifying the effectiveness in capturing complex space-time interactions.

The vehicle control method, device and vehicle provided by the present disclosure relate to intelligent driving technology, and include the following steps: determining a first scene corresponding to a vehicle and influencing a road shape, and a first path corresponding to the first scene, according to acquired external environment information of the vehicle, global path planning information, and a preset matching mode; determining a second scene on the first path and influencing a road speed, and a position of the second scene, according to a preset map; updating the first path according to the position of the second scene and the acquired position of the vehicle; processing the updated first path according to a preset motion planning to obtain a target path; and controlling the driving of the vehicle according to the target path. The present scheme can take scene analysis as a part of decision-making, so that the scene analysis and the planning module have unified judgment standards, and thus the success rate of scene switching can be improved.

The invention discloses a traffic influence autonomous evaluation system based on natural language processing and a working method, and belongs to the technical field of traffic management. In order to solve the problem of difficulty in urban traffic multi-scene analysis iteration, the system comprises a basic data layer, an NLU engine / intelligent processing layer and a user interaction layer, and the basic data layer, the NLU engine / intelligent processing layer and the user interaction layer are connected in sequence; the basic data layer comprises a road network facility database, a model parameter library, a traffic demand matrix and a traffic partition database; the NLU engine / intelligent processing layer comprises a natural language understanding NLU engine, a traffic analysis modeling evaluation module and a report generation module; and the user interaction layer comprises a Web / App graphical user interface, a voice / text input interface, a parameter confirmation and modification panel and an AR / Web visual report board, and realizes interaction between a user and the traffic influence autonomous evaluation system based on natural language processing. The method has the high efficiency of the whole process.

The specification discloses a video scene recognition method and device, a storage medium and an electronic device, wherein the method comprises: acquiring a target video frame of a target video, inputting the target video frame into a scene analysis network, determining a first region image feature corresponding to the target video frame and a second region image feature corresponding to a reference video frame through the scene analysis network, and performing time sequence feature fusion on the first region image feature and the second region image feature through the scene analysis network to output a region category graph corresponding to the target video frame. The specification can improve the accuracy of video scene recognition.

The invention discloses a visual identification and multi-modal fused target object identification method, and the method comprises the steps: constructing a model which comprises an image processing module, a rapid segmentation module and a multi-modal identification module, and enabling a candidate region outputted by the rapid segmentation module to enter a large model reasoning link after the candidate region is screened through a confidence threshold value; a confidence coefficient threshold value is dynamically adjusted through association adaptation of a threshold value parameter and target scene complexity in combination with real-time scene analysis, directional training is carried out by adopting a small-batch gradient descent method, a fine-tuned large model is used as a teacher model, and a lightweight student model is constructed through combined training of soft tags and hard tags. According to the method, the real-time advantage of a rapid segmentation module is reserved, the recognition precision is improved through the semantic understanding capability of the large model, the model volume is compressed to be less than 1 / 5 of the original large model on the premise of ensuring that the precision loss is less than 2%, the reasoning speed is improved by more than three times, and ineffective large model reasoning is reduced.

The invention relates to the technical field of intelligent agriculture, and discloses a scene analysis method and system based on multi-modal understanding, and the method comprises the steps: obtaining the multi-modal perception data of fruits, and generating an initial state data set; performing texture analysis based on the surface image, calculating roughness and uniformity parameters, and combining to obtain a skin toughness index; analyzing near-infrared reflectivity characteristics, calculating a sugar degree index through a regression model, calculating moisture content through a correction model, and evaluating the maturity; based on the toughness index and the maturity value, initial picking parameters including the speed, the clamping force and the angle are generated; collecting contact force data in real time, and calculating deformation quantity and deformation rate indexes; comparing the deformation rate with a damage threshold value to generate an adaptive force control sequence; and recording a picking result, and optimizing a parameter mapping rule. The technical problem that traditional fixed-mode picking cannot adapt to fruit individual differences is solved.

The application discloses an adaptive sound effect management method and system with a scene analysis module, relates to the field of adaptive sound effect management, and comprises the following steps: S1, collecting environmental acoustic characteristics and user behavior data in real time through a scene analysis module; S2, dynamically extracting user hearing preference parameters and scene classification labels; S3, constructing a virtual listening sweet spot matched with a physical space; and S4, dynamically adjusting the phase delay and energy distribution ratio of each frequency band to generate an optimal loudness equalization curve meeting multi-scene constraints. Through a multi-level perception-optimization closed-loop architecture, the application realizes adaptive matching of sound field characteristics in a complex dynamic scene, deeply integrates acoustic characteristic extraction and physical space modeling, breaks through the technical boundary of traditional pure signal processing, and solves the multi-objective conflict problem based on cross-modal collaborative learning and a global optimization algorithm.

This application relates to the field of data processing technology and provides a method and apparatus for acquiring trajectory scene data. In this method, trajectory data of a target vehicle is first acquired, with sources including real roadside trajectory data and simulated trajectory data. Based on the real roadside trajectory data and simulated trajectory data, the vehicle's driving intention is determined. Based on the real roadside trajectory data, simulated trajectory data, and the vehicle's driving intention, it is determined whether the target vehicle has met preset scene triggering conditions. If the target vehicle meets the scene triggering conditions, the trajectory data of the target vehicle and its corresponding driving intention within a preset time period before and after the current moment are saved as trajectory scene data. Thus, simulated trajectory data is generated through simulation, and further combined with real roadside trajectory data and simulated trajectory data for scene analysis, ultimately saving the trajectory scene data. This ensures the validity of the trajectory scene data while reducing the cost of acquiring it.

The invention discloses a method for detecting the processing quality of a remote sensing image, and belongs to the technical field of remote sensing image detection, and the method comprises the steps: obtaining a to-be-detected remote sensing image, carrying out the preprocessing and scene analysis of the to-be-detected remote sensing image, recognizing an application scene, constructing a multi-dimensional quality index according to an image index, and carrying out the normalization processing. Matching a scene evaluation template according to the normalized multi-dimensional quality indexes, calling the multi-dimensional quality indexes in the corresponding template according to an application scene of a to-be-inspected remote sensing image, performing feature fusion on the to-be-inspected remote sensing image based on the multi-dimensional quality indexes, and establishing a defect inspection model; the to-be-detected remote sensing image after feature fusion is analyzed, a defect detection result is output, a quality evaluation result is output based on the defect detection result, an overall quality score is output, defect coordinates and correction parameters can be provided, a direct basis is provided for subsequent image optimization, and the actual application value of a detection result is improved.

The invention discloses a multi-area linkage fresh air handling unit cooperative control system and method, and relates to the technical field of building environment control, and the auxiliary method comprises the specific steps that a multi-source data real-time collection module collects and transmits multiple types of data through distributed equipment; the data fusion processing module cleans and fuses the data to form a structured data set; the intelligent scene analysis and game modeling module identifies scenes, evaluates requirements, detects conflicts and quantifies utility values; the control mode decision module determines an optimal mode and parameters; the execution regulation and control module adjusts parameters, executes operation and feeds back data to form a closed loop; a multi-source data acquisition fusion framework is constructed, traditional limitation is broken through, an intelligent module automatically recognizes scenes, quantifies requirements, detects conflicts and optimizes resource allocation, meanwhile, full-process automation is achieved through closed-loop design, control economy is ensured through an algorithm, airflow interaction is optimized, energy consumption is reduced, energy efficiency and comfort are improved, and the method is suitable for large-scale popularization and application. And technical support is provided for efficient operation of green buildings.

The invention relates to the technical field of batteries, and discloses a battery test data processing method and device, equipment, a medium and a program product. The method comprises the following steps: based on battery test data corresponding to a target test item, generating a plurality of data tables with an association relationship; based on each feature dimension in the battery test data, a data label corresponding to each feature dimension is determined, and the data labels are used for representing the continuity of the corresponding feature dimensions; establishing a data bin model corresponding to the target test item based on the plurality of data tables; and on the basis of each feature dimension and the data label corresponding to each feature dimension, scene analysis data corresponding to the target test item is generated, and the scene analysis data represents chart types correspondingly generated by different data analysis scenes. According to the method, redundant data can be reduced, an efficient and convenient data query service is provided, the battery test data analysis and chart generation efficiency is improved, the battery research and development cost is reduced, and the analysis difficulty of massive battery test data is reduced.

The application particularly relates to a power system risk assessment method based on a primary-secondary coupling multiple fault model, which comprises the following steps: calculating a power transmission line fault probability of a power system; performing multiple fault sampling based on a Latin hypercube sampling method; establishing a corresponding primary-secondary coupling multiple fault model for each scene; determining a power transmission line fault state and a protection action state, and then assessing a risk index of the corresponding scene; calculating a corresponding comprehensive severity index based on the risk index of a single scene, and then obtaining one or more scenes with the highest comprehensive severity index through screening and sorting to analyze weak links of the power system; and calculating a system risk value by comprehensively considering the risk index of all scenes and a protection refusal probability, so as to assess the overall operation risk of the power system. The application can reflect the influence of the expansion of the fault range caused by incorrect protection action on the operation safety of the power grid, and can normalize and scientifically sort different dimension risk indexes.

The application relates to a kind of artificial intelligence-based dam-break scene unmanned aerial vehicle monitoring method and system.The method comprises: obtaining hydrological data, meteorological data, satellite image data and topographic map data, and inputting into monitoring area identification artificial intelligence model, generating target monitoring area risk level information and target monitoring area detection range information;Based on target monitoring area risk level information and target monitoring area detection range information, the number of unmanned aerial vehicle and the preset route of each unmanned aerial vehicle are allocated;Control each unmanned aerial vehicle to fly along the corresponding preset route, obtain the target monitoring data of each unmanned aerial vehicle;The target monitoring data of each unmanned aerial vehicle is input into dam-break scene analysis artificial intelligence model, and dam-break scene analysis result is generated.The method can realize the efficiency, accuracy and real-time of dam-break monitoring through multi-source data driving, dynamic task allocation and artificial intelligence model optimization.

A traffic accident scene analysis system comprises a photographing unit, an input unit, a user interface and an arithmetic unit electrically connected with the photographing unit, the input unit and the user interface, the photographing unit photographs a scene image of an accident to output a scene picture, the input unit is used for transmitting record information, and the arithmetic unit is used for transmitting the record information to the user interface. The arithmetic unit is used for carrying out image segmentation on the scene picture to generate at least one accident vehicle information and one road information, carrying out word segmentation on the record information to generate a plurality of pieces of keyword information, and executing program data of a generative artificial intelligent model; and generating a preliminary analysis report containing an accident animation according to the at least one piece of accident vehicle information, the road information and the plurality of pieces of keyword information, and displaying the preliminary analysis report through the user interface.