554 results about "Scene matching" patented technology

The invention relates to a sensitive data exception cross-border detection method and system based on flow analysis, and the method comprises the steps: obtaining a plurality of to-be-detected target flow sessions; analyzing each target flow session through a dynamic risk assessment model, and performing risk assessment based on the risk score of the report information, the risk score based on the historical behavior and a comprehensive risk assessment rule matched according to the current service scene to obtain a risk value; the dynamic risk assessment model is dynamically adjusted according to data flow behavior; and when it is detected that the risk value corresponding to the target stream session exceeds a preset risk threshold, triggering real-time early warning. Through application of the method and the device, the problem that sensitive data is misreported or misreported due to the fact that an existing detection method is difficult to dynamically adapt to the change of the data flow in a complex scene of cross-border data flow is solved, the change of the data flow is dynamically adapted to the complex scene of cross-border data flow, and the sensitive data is prevented from being misreported or misreported.

The invention relates to an unmanned aerial vehicle multi-modal visual positioning method and system in a GNSS denial environment, and the method comprises the steps: constructing a visual inertial mileage calculation method based on deep learning feature extraction and matching, calculating the relative positioning of an unmanned aerial vehicle, and outputting to obtain high-frequency 6-DoF pose information; performing scene matching of infrared RGB image multi-modal fusion based on the Officient-LoFTR, and obtaining absolute positioning in a world system by solving a relative pose between a current key frame picture and a matched satellite image; a factor graph optimization mode is adopted to carry out self-adaptive weight information fusion according to the confidence coefficient of absolute positioning and relative positioning under a world system, and 6-DoF pose information with high frequency and higher precision is output; based on high-frequency and higher-precision 6-DoF pose information, after the unmanned aerial vehicle is recovered, all aerial images and satellite images are fused and updated. The system comprises a calculation module and a fusion updating module. According to the invention, more accurate and high-frequency pose output can be provided.

The invention, which relates to the technical field of intelligent interaction, discloses an AI-based adaptive scene intelligent interaction system comprising a multi-modal data acquisition module, a modal preprocessing module, a multi-modal embedded coding module, an intention fusion and representation module, a service scene matching module and a service execution and reinforcement learning module. The method comprises the following steps: acquiring multi-modal original data in a user interaction process, including voice signals, text input and user behavior tracks, and synchronously recording an acquisition timestamp; according to the method, through a multi-modal unified embedding and dynamic weighting mechanism, the problem of characteristic dimension imbalance is effectively solved, and the user intention recognition accuracy is improved; meanwhile, reinforcement learning and a multi-factor scoring model are combined, personalized scene matching and dynamic response are achieved, the adaptive capacity and service accuracy of the system in a complex environment are improved, and therefore the stability and user experience of the intelligent interaction system are remarkably optimized.

The invention relates to the technical field of virtual power plant regulation and control, and provides a virtual power plant multi-scene cooperative regulation and control method and system for power supply insurance, and the method comprises the steps: periodically obtaining the supply and demand time series data of a virtual power plant, and carrying out the supply and demand fluctuation characteristic analysis of the supply and demand time series data of the virtual power plant, and generating supply and demand fluctuation time series data; according to the supply and demand fluctuation time sequence data, scene matching analysis is carried out based on a pre-constructed multi-scene regulation and control strategy library to obtain an initial scene regulation and control strategy; performing scene adaptability simulation analysis on the initial scene regulation and control strategy, and optimizing the initial scene regulation and control strategy according to a simulation analysis result to obtain a target scene regulation and control strategy; and performing layered and partitioned disassembly on a regulation target in the target scene regulation strategy to obtain a layered and partitioned regulation target, and executing virtual power plant regulation according to the layered and partitioned regulation target. According to the method, the collaborative optimization capability and the operation toughness of the virtual power plant in a complex and uncertain environment can be remarkably improved, and reliable technical support is provided for power supply insurance.

The invention relates to the technical field of artificial intelligence, in particular to a scene application operation maintenance platform based on an artificial intelligence model, which comprises a reasoning state sensing module, a load balancing scheduling module, an intelligent resource allocation module, a deployment configuration adjustment module and a scene feedback integration module. According to the method, by sensing and reasoning process resource consumption and response time migration trend, abnormal tasks caused by resource fluctuation can be accurately recognized, the state monitoring sensitivity is improved, task allocation records are analyzed to construct conflict structures, the resource conflict recognition and visualization capability is enhanced, the model operation level and the GPU vacancy rate are combined, a task migration path is optimized, and the task migration efficiency is improved. Node resource balance is realized, storage frequency, memory margin and network delay are fused during deployment, migration stability is guaranteed, drift tasks are identified through response logs and call records, an operation and maintenance list is generated, operation and maintenance accuracy and multi-scene matching capability are improved, and reasoning task stability and resource scheduling intelligent level are enhanced.

The invention relates to the technical field of intelligent scene recognition, and discloses an intelligent scene recognition method and equipment for an intelligent interaction control unit, and the method comprises the steps: carrying out the collection and preprocessing of an operation sequence, video content and audio data of the intelligent interaction control unit in a digital set-top box, and obtaining a standard multi-mode data set; inputting the standard multi-modal data set into a multi-branch deep neural network model, and extracting time sequence, visual and audio features through parallel processing to obtain a fusion scene feature vector; carrying out dimensionality reduction and semantic mapping on the fused scene feature vector to obtain a compressed semantic feature vector; performing similarity calculation on the compressed semantic feature vector and a predefined scene query vector to obtain a scene matching score; and based on the scene matching score, integrating the output of the plurality of sub-classifiers to obtain a target scene classification result, and the method can be efficiently operated on the digital set top box with limited resources to realize real-time scene recognition.

The invention belongs to the technical field of vehicle control, and particularly relates to a vehicle-mounted image recognition and target detection system based on deep learning, and the system comprises a distributed monitoring module which collects the operation, obstacle and traffic signal information of a target vehicle through multi-modal classification and scene matching, completes the marking of a shielding region and the matching of information through the combination of shared data, and achieves the recognition of the target vehicle. Forming an enhanced monitoring set; the label planning module constructs an enhanced topological space based on the enhanced monitoring set, and adjusts moving tracks in different scenes by combining with vehicle and pedestrian track probability distribution fed back by dynamic intention recognition; the action recognition module predicts trajectory parameters and collision probabilities of non-target vehicles and pedestrians by using Bayesian and multi-modal algorithms; the decision-making module generates a real-time control instruction through particle swarm optimization and fuzzy control, and optimal control parameters are fed back through simulation; according to the invention, intelligent track planning and real-time control in a complex scene are realized, and the detection precision and control robustness of the shielded and label-free area are improved.

The invention discloses a 6G digital twinning online channel modeling method, which comprises the following steps of: based on information acquired from an environment data source, carrying out communication scene online identification by utilizing a machine learning model; matching channel parameters of a 6G universal channel model according to the identified communication scene, and mapping the channel parameters to the communication scene; reconstructing the digital twin online channel of the static scene in combination with the obtained channel parameters; dynamically sensing the physical environment, acquiring environment change information, and obtaining dynamic sensing data; and in combination with a static-dynamic mixed channel modeling algorithm and dynamic sensing data, dynamically updating the digital twin online channel of the reconstructed static scene to obtain a new real-time twin channel, and constructing a visual platform of the digital twin online channel. According to the method, the channel characteristics of the existing network communication environment can be reflected in real time, real-time sensing, prediction and decision making can be realized, the pilot frequency overhead is reduced, and the system performance is improved.

The invention discloses a power market large language model evaluation method and device based on dynamic scene perception, and the method comprises the steps: building a large model evaluation system covering the three dimensions of understanding, generation and safety according to a power market application scene; the method comprises the following steps: constructing a test question bank, evaluating a power market large language model according to various evaluation methods, endowing three dimensions with different weights based on a large model evaluation system in the evaluation process, constructing a test question-scene label fine tuning data set, and performing scene classification fine tuning on the model by using a fine tuning method. Evaluating scene classification performance of the fine-tuned model, and distributing corresponding initial evaluation dimension weights for different scenes; according to the requirements and differences of different scenes, the initial evaluation dimension weight is adjusted, the scene is judged to be matched with the corresponding evaluation dimension weight according to the input content, weighted summation is carried out from the three dimensions of understanding, generation and safety, and an evaluation result is obtained. According to the method, accurate evaluation of the performance and reliability of the large language model in the power field can be realized.

The invention discloses a gamepad energy-saving control method and system based on intelligent analysis, and the method comprises the steps: obtaining the operation monitoring information of a gamepad, and carrying out the preprocessing of the operation monitoring information of the gamepad, and obtaining the multi-mode fusion features of a target gamepad; constructing a user operation analysis model, and analyzing the operation behavior of the target user using the gamepad to obtain user operation analysis information; constructing an operation state transition matrix based on the multi-modal fusion features of the target gamepad, and predicting the operation intention of the target user; constructing an energy-saving control strategy database, acquiring energy-saving control parameters matched with the current operation scene from the energy-saving control strategy database, and establishing a dynamic regulation and control mechanism to dynamically regulate and control the energy-saving control parameters; and if the matched energy-saving control parameter combination does not exist in the energy-saving control strategy database, making an energy-saving control strategy. Intelligent energy-saving control of dynamic power consumption regulation and control is realized, so that the energy efficiency performance of the gamepad and the user experience are improved.

The invention relates to the field of Internet of Things security, in particular to a network security policy generation method and system based on an AI large model. According to the method, firstly, a scene library containing historical security events and a strategy library containing various strategies are included, a to-be-processed security event currently received by a network is obtained at the same time, scene matching analysis is carried out on the to-be-processed security event and the historical security events based on data information of all dimensions, and the security event matching result is obtained. And screening out similar security events of the to-be-processed security events from all the historical security events, carrying out iterative training on the historical security events by using the constructed weak classifier about the target dimension to obtain the key degree of the target dimension, and obtaining the security events of the to-be-processed security events based on the screened key dimensions. And further performing scene matching analysis on the to-be-processed security event and the similar security event to generate a strategy set of the to-be-processed security event. According to the invention, the interference of redundant dimensions on network security event scene matching can be eliminated, and the security policy generation effect is improved.

The invention relates to the technical field of artificial intelligence, and provides a marketing data generation method and device based on portrait data, equipment and a medium, marketing association data can be collected and purified based on a three-level data gateway, feature fusion is carried out by using a star-shaped collaborative network constructed based on a dynamic weight mechanism and a federated learning mechanism, and the marketing data generation efficiency is improved. The problems of data dimension limitation and data island are solved; scene recognition is performed based on a marketing data graph constructed by a secondary scene classification tree including a gift scene, and the problems of low utilization efficiency of unstructured data and insufficient crowd portrait granularity are solved; the marketing strategy is generated by using the target engine matched with the scene, so that the problems of scene engine deficiency and gift scene adaptation imbalance are solved; and generating the target marketing data according to the target marketing strategy and the marketing data graph. The problems of low operation efficiency and insufficient content accuracy are solved.

The invention discloses an unmanned aerial vehicle-satellite scene matching method based on two stages. Precise scene matching and positioning are completed from coarse granularity to fine granularity. An existing satellite base map retrieval task is used as a first-stage task, and a local feature point matching method is introduced in a second stage. Compared with an existing method, the method has the advantages that the matching robustness in a complex scene is remarkably improved, meter-scale accurate positioning of the unmanned aerial vehicle can be realized on an existing satellite map, and a new universal solution is provided for the field of scene matching.

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of financial science and technology, medical health and the like, and discloses a scene simulation generation method which comprises the following steps: collecting text data, user data and environment data, forming a training data set, and constructing a scene simulation model based on the training data set; extracting target key information from the target text data, generating a target user portrait in combination with the target user data, and generating a scene matching strategy according to the target user portrait and the target key information; and inputting the target key information, the target user portrait and the scene matching strategy into a scene simulation model to generate a target scene simulation scheme. The text data, the user data and the environment data are integrated and labeled and classified, the scene matching strategy is generated in combination with the user portrait and the target key information, the scene simulation scheme conforming to the user characteristics is generated by utilizing the scene simulation model, accurate matching of terms and user requirements is realized, and the user experience is improved. And the user can understand the application range and effect of the clauses.

The invention relates to the technical field of radar data processing, and discloses a scene matching guidance system based on a miniature synthetic aperture radar, which comprises a multi-dimensional acquisition module and an intelligent matching module. According to the system, SAR image data of all target areas and SAR image data provided by echo signals in real time are acquired through a multi-dimensional acquisition module and are classified to form a data set, the multi-dimensional acquisition module performs standardization processing and analysis on the data set to generate a digital elevation model, an intelligent matching module analyzes the authenticity of SAR images of each target area, and the authenticity of the SAR images of each target area is analyzed. The method comprises the following steps: generating a corresponding credible index, setting a monitoring period with a fixed time length, analyzing a fluctuation index of real-time signal intensity and a matching index of an SAR image, quantifying data quality and matching potential, dynamically monitoring high image precision, setting a fluctuation threshold value by an intelligent matching module, judging whether the SAR image has a delay problem or not, and judging whether the SAR image has a delay problem or not. And scene matching guidance and track synchronization are carried out according to the credibility index, the matching index and the digital elevation model, and the scene matching guidance is high in applicability.

The invention belongs to the field of road vehicle control systems, and relates to a vehicle trafficability risk judgment method and system fusing driver ability and confidence. The system comprises a driving behavior and environment data acquisition module, a scene matching module, a traffic capacity estimation module, a traffic capacity scoring module, a driving behavior feature analysis module, a traffic confidence scoring module, a trafficability risk calculation module and a database. The scene matching module is used for carrying out similarity calculation on the basis of a historical traffic sample set and the structural features of the current traffic scene and screening similar samples; the traffic capacity estimation module is used for carrying out weighted fusion on the screened traffic capacity and the calculated matching weight to obtain a traffic capacity estimation result; and the trafficability risk calculation module is used for correcting the original trafficability risk based on the structure complexity and the vehicle performance by using the traffic capacity score and the traffic confidence score. According to the system, personalized and dynamic adjustment of traffic risks is realized, and the accuracy of trafficability risk prediction is improved.

The invention discloses a vibration space-time diagram event identification method based on distributed optical fiber sensing. According to the invention, through combination of spatial-temporal feature fusion and a deep learning technology, the accuracy and robustness of event identification are significantly improved. The method comprises the following steps of: firstly, integrating space-time dimension information of an optical fiber vibration signal into two-dimensional characteristics by a space-time diagram, and providing richer context association for a model; through introduction of the channel attention and space-time attention module, key features can be adaptively focused, noise interference can be suppressed, and the adaptability of the model to a complex background is enhanced. Meanwhile, the dynamic loss function optimizes the training process of different scale targets by dynamically adjusting the loss weights of the bounding box and the mask, and solves the problem of performance reduction caused by target scale difference in the traditional method. The method has strong generalization, can adapt to various optical fiber types and monitoring scenes through data enhancement and adaptive parameter adjustment, and reduces dependence on data annotation and scene matching in actual deployment.

The invention relates to the technical field of scene matching navigation, and provides a multi-mode scene matching navigation positioning method and device for a self-adaptive variable-structure unmanned aerial vehicle. According to the method, geographic coordinates corresponding to all pixel points in an aerial image are determined according to pose information in the aerial image of the unmanned aerial vehicle, and the geographic position of the unmanned aerial vehicle with errors is corrected based on difference information between the pixel coordinates of the central point of the aerial image of the unmanned aerial vehicle and the geographic position corresponding to the center of a camera. Based on the corrected geographic position of the unmanned aerial vehicle, determining a reference image pixel coordinate corresponding to the angular point pixel coordinate of the aerial image, and obtaining a simulated aerial image under the view angle of the unmanned aerial vehicle through perspective transformation so as to determine a homography transformation matrix; according to the method, the satellite reference image pixel coordinates of the interest point in the aerial image are determined based on the homography transformation matrix and the perspective transformation matrix, and finally the longitude and latitude of the interest point are determined, so that the limitation of the traditional method in view angle difference compensation, computing resource constraint and cross-modal processing is solved, and the universality, robustness and positioning accuracy of the system are improved.

The embodiment of the invention provides a three-dimensional Gaussian scene reconstruction method and device, electronic equipment and a storage medium, and the method comprises the steps: firstly, obtaining a plurality of three-dimensional Gaussian distributions of a target scene; then, obtaining a two-dimensional projection weight of each three-dimensional Gaussian distribution, and selecting a key Gaussian from the plurality of three-dimensional Gaussian distributions based on the two-dimensional projection weights; next, performing Gaussian splitting on the key Gaussian to obtain a plurality of split Gaussian distributions, and performing depth re-initialization based on the plurality of split Gaussian distributions and the plurality of three-dimensional Gaussian distributions to obtain a plurality of dense three-dimensional Gaussian distributions; and finally, selecting a plurality of reserved Gaussian distributions from the dense three-dimensional Gaussian distributions based on the two-dimensional projection weights of the dense three-dimensional Gaussian distributions, and performing scene reconstruction based on the reserved Gaussian distributions to obtain a reconstructed scene matched with the target scene. Therefore, the overall efficiency of three-dimensional scene reconstruction and the compactness and quality of the finally reconstructed scene are obviously improved.

The invention provides a multi-round dialogue question and answer method and system, electronic equipment and a storage medium, and the method comprises the steps: obtaining a current question, historical dialogue information and / or a user portrait of a user; obtaining a plurality of predefined scenes in the scene information configuration file, performing scene matching on the current problem, and determining predefined information of a target scene corresponding to the current problem; on the basis of a pre-constructed inheritance model, according to the predefined information of the target scene, whether the historical dialogue information and / or the user portrait need to be inherited or not is judged, if yes, the current question is complemented in combination with the predefined information of the target scene, the historical dialogue information and / or the user portrait, and the complemented current question is output; wherein the inheritance model is obtained by training based on the sample question, the sample predefined information of the corresponding sample scene, the sample dialogue information and / or the sample user portrait. The scene adaptability is high, the model training efficiency is improved, and the model training cost is reduced.

The embodiment of the invention provides a network management method and device, and relates to the technical field of communication, the network management method is applied to a first node for information prediction in a communication network, and comprises the following steps: obtaining reference data matched with a network management scene and current measurement information, the reference data is used for indicating the prediction quality of at least one prediction model matched with the scene; processing the current measurement information through each prediction model in the at least one prediction model to obtain current prediction information corresponding to the prediction model; and sending the current prediction information corresponding to each prediction model and the quality index of the current prediction information to a second node in the communication network, so that the second node performs network management operation aiming at the scene according to the current prediction information of which the quality index meets the quality condition in the current prediction information corresponding to each prediction model. According to the embodiment of the invention, the prediction information can be subjected to quality evaluation, so that the network performance and the user experience are improved.

The invention provides a multi-modal feature vector fused vehicle cross-lens ReID matching method and system, and relates to the technical field of vehicle identification, and the method comprises the steps: collecting a vehicle image through a camera, and extracting triple features, including a bottom-layer visual feature, a middle-layer semantic feature and a high-layer character feature; in combination with vehicle space-time metadata at adjacent acquisition cameras, generating space-time trajectory feature vectors, and capturing a time sequence rule of a vehicle driving path; the missing features are complemented by adopting an AI generation technology, and the complementation precision is optimized through adaptive learning; inputting the triple feature, the time sequence rule and the complementation feature into a Transform fusion network to generate a fusion feature vector; calculating the cosine similarity of the to-be-matched vehicle according to the fusion feature vector, and carrying out cross-border head matching; according to the method, the problem of insufficient distinction degree of single-modal features is solved, the track feature vector is generated by combining the space-time metadata of the camera, the driving time sequence rule of the vehicle between different lenses is captured, and the space-time constraint of cross-scene matching is enhanced.

The invention provides an efficient data retrieval method and system oriented to a large language model. The method comprises the following steps: determining an intention vector set matched with a user intention; determining a specified number of candidate text fragments with the highest correlation with the user intention in an index database; obtaining multi-dimensional feature evaluation data of each candidate text fragment, and constructing a multi-dimensional feature numerical matrix of each candidate text fragment according to the multi-dimensional feature evaluation data; according to the multi-dimensional feature numerical value matrix, selecting selected text fragments from the candidate text fragments; sorting the selected text fragments, and sequentially performing scene matching on the selected text fragments and a pre-cached high-frequency query scene according to the sorting, so as to extract cache content matched with the target high-frequency query scene from the cache as a query result when a matching result represents that the current query scene accords with the target high-frequency query scene. The invention provides an efficient data retrieval method, and the quality and precision of data retrieval are improved.

The invention relates to the technical field of intelligent light regulation and control, and discloses an intelligent light regulation and control method based on scene matching, and the method comprises the following steps: presetting light modes, and setting light parameters for each light mode; collecting environmental parameters, counting historical light data, and calculating a recommendation index of each light mode based on the environmental parameters and the historical light data; based on the recommendation index, calculating a comprehensive recommendation index of each light mode, and generating a recommendation table of the light modes; setting the light mode as the light mode with the highest comprehensive recommendation index in the recommendation table; collecting activity intensity indexes of the user, and calculating the matching degree between each light mode in the recommendation table and the current scene; and setting the light mode as the light mode with the highest matching degree with the current scene in the recommendation table. According to the invention, accurate matching and personalized customization of the light mode are realized, the user experience is improved, and the energy consumption and the illumination cost are reduced at the same time.

The present application relates to a machine tool processing method, device, computer equipment, computer-readable storage medium and computer program product, and relates to the field of machine tool processing technology. The method includes: in response to a processing instruction for a target machine tool, determining the error compensation amount adapted to the current processing scene of the target machine tool through the error compensation model of the target machine tool; the error compensation model is a model constructed based on error compensation parameters matching the deviation data of the processing scene for each processing scene matching the target machine tool; generating an error compensation instruction matching the error compensation amount; according to the error compensation instruction, performing error compensation on the processing instruction, generating an actual processing instruction for the target machine tool; executing the actual processing instruction, controlling the target machine tool to process, and obtaining the processing result of the target machine tool. The use of this method can ensure the processing accuracy of the machine tool under different processing scenarios.

The invention discloses a multi-scene term translation model training method and processing method based on a large model, and the method comprises the steps: firstly constructing a term-scene-translation triple data set, and formatting the term-scene-translation triple data set into a training sample containing an instruction, context input and expected output through a structured term translation instruction set; then, the samples are used for fine tuning of the basic large language model, and a preliminary adaptive model is obtained. And finally, introducing a reinforcement learning method based on a verifiable reward mechanism to optimize the fine tuning model, and iteratively improving the performance of the model in the aspects of scene matching degree, specification conformity, cross-scene consistency and the like by calculating a reward value. According to the method, the problems that a traditional term translation model is poor in flexibility and insufficient in specialty are effectively solved, and the multi-scene term translation which is high in accuracy and conforms to specific industry specifications can be automatically generated.

The invention relates to the technical field of dangerous cargo management, in particular to an intelligent compliance verification system for dangerous cargo boxing operation. Comprising a data acquisition preprocessing unit; the law and regulation graph construction iteration unit is used for extracting a'law and regulation term number-verification dimension-constraint condition 'triple of the law and regulation in the dangerous cargo boxing field and introducing an improved three-dimensional dynamic weight conflict resolution algorithm according to the efficacy level, the release subject and the application range indexing; a multi-agent collaborative verification unit; a risk early warning report generation unit; and a man-machine cooperation closed loop unit. According to the method, the problems of lagging of laws and regulations and clause conflicts are solved by adopting triple extraction of'laws and regulations clause numbers-verification dimensions-constraint conditions', a three-dimensional dynamic weight conflict resolution algorithm based on laws and regulations effectiveness-scene matching degree-release time and a technical means of periodically optimizing weight parameters through a reinforcement learning model.

The invention discloses a method for monitoring and identifying dangerous behaviors of a smart park, and relates to the field of intelligent security and protection. The method comprises the following steps: acquiring an environment data stream through a sensing device deployed at a key point location of a park and carrying out standardization processing on the environment data stream; the data stream is sent to an edge computing node, a target is recognized through a lightweight target detection model, the detection confidence coefficient is obtained, a motion track is generated through a multi-target tracking algorithm, the rule matching degree is calculated according to a predefined rule base, and dangerous behaviors are preliminarily recognized; calculating a comprehensive model prediction credibility based on the detection credibility and the rule matching degree, and generating a final comprehensive credibility through a multi-dimensional credibility fusion model in combination with the time sequence consistency credibility and the scene matching credibility; and dangerous behavior trend prediction is carried out according to the comprehensive confidence, and an alarm instruction is generated to trigger disposal linkage, so that advanced prevention and quick response of dangerous behaviors are realized, and the intelligent level and operation and maintenance efficiency of park security and protection are greatly improved.

The invention provides a business travel data management method and system based on cloud computing and scene matching and computing equipment. The method comprises the following steps: acquiring at least one geographic image uploaded by a user in a business travel plan execution process; generating a compliance geographic region set according to the travel plan; matching the geographic image with the compliance geographic region set to obtain a first matching result; identifying and extracting visual scene features in the geographic image, and performing consistency cross validation comparison on the visual scene features and a geographic scene expected based on geographic coordinates to obtain a first comparison result; generating an authenticity verification result of the travel behavior based on the first matching result and the first comparison result; and when the authenticity verification result is that the verification is passed, automatically generating an electronic audit voucher containing a spatial trajectory chain and a geographic image. According to the technical scheme of the invention, the accuracy of authenticity verification of travel behaviors can be improved, a high-credibility electronic audit voucher is automatically generated, and a data closed loop is provided for intelligent travel management.

The invention discloses an intelligent broadcasting director system for a dance scene, and the system comprises a pull stream decoding module which is used for obtaining video streams from a plurality of cameras and decoding the video streams; the fuzzy detection module is used for evaluating whether each frame of image in the video stream is clear or not; the dance formation change judgment module is used for analyzing the decoded video frame based on a target tracking algorithm and detecting and judging whether the formation of the dance actor changes or not; the leader dance recognition module is used for recognizing leader dance actors in the decoded video frames through a deep learning human body posture recognition algorithm; the scene switching module is used for obtaining a scene-based shot transition probability matrix and a shot playing duration matrix according to a statistical result of the dance program video data, and is used for guiding the staying duration of a shot and switching channel selection according to scene matching when no formation change exists and the leader dance does not have special actions; and the intelligent director logic switching module is used for automatically executing a video switching instruction according to the demand of the dance scene so as to realize an intelligent director effect. According to the invention, intelligent broadcasting of a dance scene can be realized, and the efficiency and quality of program production are improved.