2402 results about "Behavior recognition" patented technology

The invention relates to the technical field of data center management, and discloses a data center digital twinborn simulation and decision-making system oriented to intelligent management. The system comprises a data center physical feature sensing module, a virtual space reconstruction module, an operation situation deduction engine, an abnormal behavior recognition module and a decision instruction generation module. Wherein the physical feature sensing module collects multi-dimensional operation parameters of the infrastructure in real time; the virtual space reconstruction module dynamically constructs a three-dimensional virtual model based on the collected parameters; running a situation deduction engine to simulate a resource scheduling and energy flow process; the abnormal behavior recognition module analyzes the analog data stream to detect an abnormal operation mode; and the decision instruction generation module integrates the abnormal information and generates an optimization regulation and control instruction for the physical equipment. According to the system, intelligent management of the data center is realized through real-time mapping, dynamic deduction and intelligent decision making of physical and virtual spaces, and the management precision and timeliness are improved.

The invention discloses a power transmission line multi-mode warning system and a power transmission line multi-mode expelling method, belongs to the technical field of power transmission line safety monitoring, and aims at solving the problems that power transmission line invasion target monitoring is not accurate, and the warning and expelling effect is poor. An environment image is acquired through an acquisition unit, and redundancy compensation is carried out on a fault unit. In the aspect of image splicing, a database is constructed by using structural feature points of a power transmission line, rapid projection transformation of a fixed area is realized, incremental feature matching is adopted for a dynamic area, and an environment panoramic image is generated. The method comprises the following steps: establishing a background coordinate system based on an environment panoramic image by aligning a fixed structure region, extracting multi-modal features of an intrusion target, constructing a dynamic trajectory parameter set, completing species classification and behavior recognition, constructing a multi-dimensional evaluation index system, dividing threat levels, and generating a thermodynamic diagram. And finally, according to data such as threat levels, a multi-mode grading warning system is constructed, warning equipment is dynamically adjusted, a target track is tracked, a warning effect is evaluated, and accurate and efficient invasion target expelling is realized.

The invention, which relates to the technical field of community safety supervision, discloses an artificial intelligence-based community safety environment supervision system comprising a data acquisition module, a data processing and analysis module, an intelligent decision module, an early warning response module and a system management module. The data acquisition module is used as a sensing layer of the system; the data processing and analysis module specifically comprises a feature extraction unit and a behavior recognition unit; the intelligent decision module is used for receiving the risk assessment result output by the data processing and analysis module; and the early warning response module generates the scheme according to the intelligent decision module. According to the community safety environment supervision system based on artificial intelligence, intelligent supervision of a community safety environment is realized through a complete closed loop of data acquisition, data processing, intelligent decision making, early warning response and system management; all the modules are in close cooperation, full-process automation from data collection to emergency response is ensured, and the efficiency and accuracy of community safety management are greatly improved.

The invention discloses a financial network security defense method and system based on multiple Agents and a dynamic large model. A detection Agent is deployed in an edge layer, financial network node flow data and system logs are collected in real time, time sequence features are extracted through a lightweight convolutional network, and a preliminary anomaly score is generated. And the cloud layer constructs a decision Agent, receives the feature abstract transmitted by the edge node in an encrypted manner, inputs the feature abstract into a dynamic large model for multi-modal feature fusion, and outputs defense action probability distribution. And the intelligence Agent constructs a cross-institution federated learning network. And constructing a dynamic game engine, constructing a revenue matrix based on the attack cost and the defense revenue, solving a Nash equilibrium strategy, and generating an optimal defense instruction set. And dynamically allocating detection tasks according to the threat level and the edge computing power state. According to the method, efficient acquisition and analysis are realized, the abnormal behavior recognition capability is improved, support is provided for making a defense strategy, the defense strategy is optimized, and the intelligent, automatic and efficient levels of defense are improved.

The invention provides a photovoltaic module fault detection system, method and device based on infrared image hot spot detection, and a storage medium, and belongs to the field of photovoltaic module fault detection. The system comprises a temperature difference boundary extraction module, an image sequence registration module, a hot spot track identification module, an abnormal dynamic screening module and a fault hot spot confirmation module. Pixel difference screening is carried out by setting a temperature difference threshold value, a boundary communication structure is established, frame-level displacement calculation between time sequence images is introduced to realize coordinate alignment, hot spot center points in continuous frames are extracted to form a motion path, abnormal point screening and time node labeling are carried out in combination with path displacement characteristics, and the time sequence image is obtained. According to the method, the boundary area growth rate and temperature change double factors are fused to screen fault areas, dynamic tracking and accurate detection of abnormal hot spots are achieved, the logic relevance between abnormal behavior recognition and fault judgment is enhanced, and the integrity of photovoltaic module fault information extraction and the accuracy of hot spot recognition are guaranteed.

The invention discloses a civil aviation airport security method and system based on multi-source perception and dynamic risk assessment, and the method comprises the steps: building a personnel-article-scene ternary relation model through collecting multi-source data such as video monitoring, face recognition, luggage scanning and sound signals; based on an abnormal matching mode of the model, dynamic risk assessment is carried out by combining real-time people flow density and historical event thermodynamic distribution, and a dynamic risk calibration map is generated; an identification strategy parameter is configured according to the risk level, a self-adaptive strategy set is formed and applied to the intelligent terminal, target identification and behavior tracking are achieved, and a security and protection system is automatically linked when a triggering condition is met; according to the method, the behavior-article-scene ternary relation model is constructed, so that multi-source data collaborative analysis is realized, the limitation of traditional single-mode detection is overcome, the abnormal behavior recognition accuracy in a complex scene is remarkably improved, and the false alarm rate is effectively reduced.

The invention discloses a power plant specific area personnel behavior identification and early warning system, and relates to the technical field of industrial safety intelligent monitoring, the power plant specific area personnel behavior identification and early warning system comprises a data acquisition module, an edge calculation module and a central processing platform, the edge computing module carries out lightweight processing and localized decision making on original data, the load of the central processing platform is reduced, and the central processing platform realizes full-process safety management and control of a high-risk area through multi-modal data fusion, dynamic authority management and trajectory analysis. According to the invention, lightweight AI models are embedded in various cameras of a work card authentication terminal and a visual perception unit, preliminary reasoning is directly completed at the work card authentication terminal and the cameras, so that the work card authentication terminal and the cameras have edge computing capability, and independent edge servers or industrial personal computers are deployed near data sources of various regions of a power plant. And serving as a regional edge computing node.

The invention discloses a human abnormal behavior monitoring method based on a large-model multi-agent, which is executed by a modular multi-agent system deployed on a back-end server, obtains information through a monitoring camera, and comprises the following steps: obtaining a video stream from the monitoring camera by a sensing agent and extracting human body posture features; analyzing the key frame by a scene understanding agent by using a visual large model, and constructing a time sequence dynamic scene graph; the core reasoning agent evaluates the scene semantic conformity based on the pre-trained large model and performs abnormal preliminary judgment; performing fine-grained classification, interpretation generation and risk assessment on the abnormal behaviors; and the report and action agent generates an alarm and records event data. According to the invention, through multi-agent cooperative work and a large model technology, efficient and accurate monitoring of human abnormal behaviors is realized, and the intelligent level of the monitoring system and the abnormal behavior identification accuracy are improved.

A dog training method for intelligently correcting pet behaviors based on multi-modality comprises the steps that multi-modality perception data of a pet is obtained, the multi-modality perception data comprises acoustic signals, motion signals and track data, an acoustic feature tensor, a motion feature tensor and a position feature tensor are constructed, and a unified coupling tensor is formed; inputting the coupling tensor into a pre-trained behavior recognition model, and outputting a behavior tag; the emotion recognition module outputs an emotion label, and updates a character label in a sliding window mode; inputting the behavior label, the emotion label and the character label into an intervention strategy generation model and outputting an adaptive intervention strategy, wherein the intervention strategy comprises a pacifying type strategy and a stimulation type strategy; according to a historical feedback effect record, adjusting a feedback intensity parameter of the intervention strategy, calculating a feedback convergence index and an emotion recovery index, and dynamically correcting a feedback level and a delay strategy of the intervention strategy.

The invention relates to the technical field of knowledge extraction, in particular to a knowledge graph construction method and system based on a large model, and the method comprises the following steps: obtaining a current input statement of a user through a dialogue state tracker, inputting the statement into a BERT intention classification model for domain label analysis, behavior type recognition and emotional tendency detection, and outputting a three-dimensional classification vector; and extracting entity lexical items and relation predicates based on an LSTM sequence tagging device, and generating an original semantic structural body. According to the method, intention classification, behavior recognition and emotion detection are fused through three-dimensional semantic analysis, semantic comprehension granularity is improved, dynamic entity disambiguation is combined with a Manhattan distance threshold value and dialogue history tracking, semantic boundaries are defined to reduce anaphora ambiguity, and cross-modal alignment is enhanced through relation predicate hierarchical clustering and knowledge base dynamic matching; generative reply and semantic coherence reordering collaboratively keep topic continuation, and structured analysis and unstructured generation closed loop optimize semantic output and interaction fluency.

The invention relates to the technical field of video behavior recognition, and discloses a behavior recognition and early warning system for a real-time video stream. The system comprises a spatio-temporal feature modeling module, a behavior fragment extraction module, an anomaly propagation modeling module, a risk area positioning module and an early warning strategy generation module. The spatial-temporal feature modeling module builds a dynamic model based on historical data, captures a skeleton key point three-dimensional coordinate sequence, a motion optical flow vector field and a micro-expression intensity spectrum, and outputs a theoretical behavior mode vector; the behavior fragment extraction module generates a multi-modal difference feature tensor through cross-modal difference analysis; the exception propagation modeling module generates an exception propagation path risk probability distribution cloud picture in combination with spatial constraint and trajectory information; the risk area positioning module identifies a high-risk area and marks a boundary; and the early warning strategy generation module dynamically configures monitoring parameters, starts high-frame-rate micro-expression capture for a high-risk area, and performs a track disturbance test on an adjacent area.

The embodiment of the invention provides a multi-modal fusion deep learning analysis method and system. The method is applied to the technical field of multi-modal learning, and comprises the following steps: obtaining multi-modal original data, sequentially processing image, text, audio and video data, and extracting visual features of the image, semantic features of the text, frequency spectrum and time sequence features of the audio, image features and time sequence features of a video frame and time domain features of an audio sequence; and then, according to the complementary information of the multi-source features, fusion processing is carried out to form a unified multi-modal feature representation, the unified multi-modal feature representation is input to a preset deep learning analysis model, and finally a multi-modal analysis result of comprehensive expression is obtained. According to the scheme, information complementarity and robustness are enhanced through multi-modal feature fusion, the comprehensive analysis capability of the model on semantic understanding, behavior recognition and state judgment in a complex scene is remarkably improved, and a more accurate, efficient and stable decision basis is provided for a multi-modal intelligent sensing system.

The invention relates to the field of image recognition, in particular to a wind power construction intelligent safety management method and system based on intelligent AI monitoring. The method comprises the following steps: obtaining an omnibearing real-time image flow of a wind power construction area, carrying out super-resolution deep convolution optimization and operator three-dimensional image segmentation, and extracting an operator three-dimensional image frame; three-dimensional point cloud modeling of the construction area is carried out based on the image flow, real-time image frame position positioning rendering is carried out according to a three-dimensional image frame, and a real-time twinborn model of the construction area is constructed; performing operation dynamic behavior analysis and behavior deviation degree quantitative analysis based on a twin model to obtain the behavior deviation degree of the operator; and according to the behavior deviation degree, carrying out early prediction analysis on illegal behaviors, making an adaptive risk early warning decision, and constructing an operation behavior risk early warning strategy. According to the invention, through real-time operation behavior identification and environmental risk analysis, the intelligence and safety level of wind power construction are improved.

The invention relates to the technical field of target detection, and discloses a pet target detection method and device and a camera, and the method comprises the steps: carrying out the motion triggering collection and image enhancement preprocessing of a front end region of a feeder, and obtaining an enhanced image frame sequence; performing feature extraction of dynamic receptive field adjustment on the enhanced image frame sequence to obtain pet feature descriptors and position information; behavior time sequence feature analysis is executed, and pet behavior sequence feature vectors are obtained; constructing a state transition diagram according to the pet behavior sequence feature vector, and performing time sequence consistency analysis to obtain a pet state judgment result; power management and decision execution are carried out on the feeder based on the pet state judgment result, feeding control under the low-power-consumption condition is achieved, behavior misjudgment caused by posture fluctuation is effectively avoided, the behavior recognition accuracy is improved, the accurate feeding control problem in a multi-pet family is solved, and the user experience is improved.

The invention relates to the technical field of state monitoring, in particular to an inner package production data real-time monitoring and processing system which comprises a running state acquisition module, a performance deviation evaluation module, an abnormal behavior recognition module, a task scheduling optimization module and a load balancing optimization module. According to the invention, through a cooperative acquisition mode of operation data such as temperature fluctuation, pressure change and flow velocity stability, real-time perception of dynamic collection and high-frequency change of the operation state is realized, and by means of a cross discrimination strategy of distribution uniformity and response time, non-representative data segments are eliminated, and the accuracy of performance abnormity identification is improved. Through time coupling analysis of gradient tracks and offset, abnormal behavior fragments are locked in advance, forward recognition of trend instability is achieved, intervention windows are dynamically screened, actual effect and synchronization of response are ensured, a real-time matching mechanism based on loads and demands is adopted, task priority and processing channel distribution are optimized, and intervention efficiency and scheduling adaptability are improved.

The invention discloses a video analysis-based multi-scene operator violation behavior identification method and system, and belongs to the technical field of intelligent operation safety monitoring and artificial intelligence identification, and the method comprises the steps: collecting a real-time video stream of a multi-scene operation site; recognizing a continuous action time sequence in the real-time video stream by using an action recognition depth model; constructing the continuous action time sequence into an action behavior sequence; the action behavior sequence is constructed into a directed behavior graph with time, space and action labels, the directed behavior graph is compared with a directed behavior graph corresponding to the standard action behavior sequence, and illegal behaviors are recognized; and carrying out multi-mode early warning on the identified illegal behaviors. According to the method, the bottleneck that the traditional image recognition technology is weak in action sequence semantic understanding and poor in environmental adaptability is broken through, and accurate recognition and real-time early warning of illegal behaviors in multi-scene operation are achieved.

The invention discloses a power distribution system flexible regulation and control system based on load side behavior recognition and a method thereof, and relates to the technical field of power distribution of power systems. The method comprises the following steps: acquiring power utilization power, equipment state, environment parameters and power utilization preference information of a user side in real time; dynamically generating and updating behavior inertia factor data, and reporting high-frequency change data when the behavior inertia factor data exceeds a preset threshold; receiving regional load feature abstract data, and uploading the abstract data when the change of the abstract data exceeds a threshold value; historical period abstract data are fused, and a prediction matrix containing the partition load trend and the peak probability in the T time window is generated through a behavior inertia dynamic evolution model; and when the load rate of the system exceeds a safety threshold value, calculating an individual adjustment amplitude, and generating a regulation and control instruction containing a time-phased target and a flexible adjustment amplitude. Accurate prediction and flexible regulation and control of the load of the power distribution system are realized, and the operation stability and the power supply quality of the system are remarkably improved.

The invention discloses a flowmeter-oriented embedded multi-parameter fusion calibration control method and system, relates to the technical field of industrial measurement and embedded control, and is used for solving the problem of unstable calibration caused by misalignment of multi-link flow signals, compensation lag and fusion oscillation. The method relates to synchronous acquisition, pretreatment and calibration of differential pressure type, ultrasonic wave and thermal type mass measurement links and environment parameters such as temperature, pressure and density. All-channel unified time scales are established through ternary binding, filtering and denoising, zero snapshot and alignment rules, a stability scoring mechanism is constructed in combination with offset behavior recognition, event classification and feature aggregation, and dynamic switching of a compensation path and a fusion strategy is driven. In a double-path fusion structure, table look-up compensation, model estimation, multi-algorithm parallelism and fusion track recording are adopted, version traceability and calibration result closed-loop control are achieved, and the stability and rechecking performance of mass flow output under complex disturbance are effectively improved.

The invention relates to an intelligent safety protection management method and system. The method comprises the following steps: acquiring behavior data and position data of personnel in an industrial site; analyzing the behavior data and the position data by using a pre-trained personnel behavior model to obtain a behavior recognition result; according to the behavior recognition result and the current operation state of the target equipment, judging a risk level corresponding to the personnel behavior; based on the risk level, a corresponding safety response strategy is matched in the edge computing node, a control instruction is generated according to the safety response strategy, and the control instruction is used for driving the target device to execute a corresponding response action; and the voice interaction module is used for acquiring voice input of an on-site operator, performing semantic recognition on the voice input to obtain a voice recognition result, and executing corresponding emergency control operation to cover a current control instruction when the voice recognition result meets a preset emergency instruction condition. The method has the effect of improving the accuracy of intelligent safety management in the industrial environment.

The invention relates to the technical field of short video encryption, and discloses a short video active defense encryption system based on a device fingerprint and a dynamic confusion field, and the key point of the technical scheme is that the system comprises a device fingerprint generation module, a mother video encryption module, a slice encryption module, a behavior recognition and defense module and an encryption logic update regulation and control module. The system generates a unique device fingerprint hash value through a multi-modal feature, generates a dynamic confusion field in combination with a chaotic system and a quantum random number, realizes differential encryption of a mother video and slices, and is embedded with zero-knowledge consanguinity proof to support traceability verification. The behavior recognition and defense module monitors user behaviors in real time and dynamically adjusts a confusion strategy or triggers an active defense mechanism, and the encryption logic updating regulation and control module optimizes the updating frequency according to playing data and reduces the batch crawling risk. According to the invention, the security and anti-attack capability of the short video content can be effectively improved.

The invention provides a multi-modal information fusion feeding decision-making system and method for chicken breeding behavior recognition, and the method comprises the steps: collecting a multi-source heterogeneous data set, and extracting a primary fusion feature vector; constructing a triple knowledge graph; mining implicit association rules of the ingestion frequency and the body temperature; performing secondary fusion on the primary fusion feature vector and an implicit association rule to generate an intermediate decision feature; and generating a feeding decision instruction through the pre-training decision model and the expert rule base. According to the method, the knowledge graph is constructed, GNN reasoning is utilized, and a manual preset rule static mode is replaced; performing secondary feature fusion, generating intermediate decision features by combining primary fusion features and implicit rules, and then combining a pre-training model and an expert rule base, ensuring decision real-time performance, integrating domain knowledge, outputting accurate adjustment parameters, realizing full-link intelligence, improving the accuracy and adaptability of breeding chicken feeding decisions, and improving the accuracy and adaptability of chicken feeding decisions. Therefore, dynamic requirements of complex breeding scenes are met, and chicken flock health and breeding efficiency improvement are promoted.

The invention discloses a security monitoring video intelligent analysis method based on multi-algorithm cooperation and unified architecture, and relates to the technical field of intelligent video monitoring, and the method comprises the steps: collecting a security monitoring video, carrying out the preprocessing, carrying out the spatial-temporal feature extraction and fusion through a CNN-LSTM spatial-temporal fusion engine, and outputting a fusion feature map; performing target detection, behavior recognition and anomaly detection on the fused feature map to obtain a multi-algorithm analysis result; a cross-module fusion mechanism based on an attention mechanism is utilized to perform weighted integration processing on the multi-algorithm analysis result to obtain a fusion event representation vector; performing event type identification and risk level evaluation on the fusion event representation vector to obtain an event classification result and an event risk level; according to the invention, through the CNN-LSTM space-time fusion engine, the front-end perception capability of abnormal behaviors in a complex scene is improved, and the event detection accuracy and the anti-interference capability are improved.

The invention discloses a video monitoring abnormal behavior identification method and system based on edge AI, and relates to the technical field of intelligent video analysis, and the method comprises the steps: carrying out the frame segmentation processing of a video monitoring data stream, obtaining a video frame sequence, building a target trajectory prediction model, predicting the position region of a current frame, and generating target position prediction data; performing multi-scale feature extraction on the video frame sequence, performing fusion matching on a target feature vector and target position prediction data, and constructing an enhanced feature matrix; carrying out weight distribution on the key behavior characteristics by adopting an attention mechanism, setting a dynamic threshold adjustment mechanism, and dynamically adjusting an abnormal behavior judgment threshold according to the personnel density; and inputting the adjusted feature data into an abnormal behavior classifier for identification and judgment, outputting an abnormal behavior identification result, and generating an abnormal event report. According to the method, the abnormal behavior detection accuracy in a complex monitoring scene is improved, the false report and missing report rate is reduced, and the millisecond-level real-time response capability is realized.

The invention provides a safety early warning method and device for abnormal behavior trajectory analysis and a medium, and relates to the technical field of safety early warning, and the method comprises the steps: collecting behavior trajectory data of a target region through a multi-source sensing device, and carrying out the multi-dimensional feature extraction of a behavior trajectory data set; according to the stay frequency characteristics, carrying out abnormity determination on the track space-time distribution characteristics and the speed change characteristics, and constructing an abnormal behavior determination vector; performing wandering mode recognition based on the abnormal behavior judgment vector, determining an abnormal behavior level, and triggering a graded early warning mechanism according to the abnormal behavior level; and generating an abnormal behavior alarm signal according to the grading early warning mechanism, combining the abnormal behavior alarm signal with the electronic fence information of the target area to generate a safety early warning report, and pushing the report to a monitoring terminal. According to the invention, the technical problem of abnormal behavior recognition accuracy in the prior art can be solved, and the technical effect of improving the behavior judgment accuracy is achieved.

The invention relates to the technical field of wind power generation, discloses an intelligent safety early warning method for wind power hoisting operation, and aims to solve the technical problems of insufficient perception and lack of data fusion and intelligent analysis decision in existing operation safety management. The method is characterized by comprising the following steps: constructing an intelligent terminal integrated with a UWB / IMU / safety belt / environment sensor; deploying a positioning base station network and a signal relay system; a multi-source data fusion intelligent analysis platform is established, and high-precision positioning, behavior recognition and deep learning risk prediction are achieved; and a closed-loop intelligent decision-making and execution system is constructed, and graded early warning, electronic fence, environment linkage and one-key help calling are realized. According to the method, a traditional experience driving mode is innovated into a data driving mode, the safety level and the operation efficiency of wind power hoisting operation are remarkably improved, and the defects of insufficient perception, information isolation and the like are overcome.

The invention relates to the technical field of multi-modal time series data processing, in particular to a multi-scale multi-modal fusion time series classification model, which comprises the following steps of: performing timestamp unification, missing value filling and normalization processing on input multi-modal time series data to generate standardized data; extracting long-period and short-period features by adopting a time sequence convolutional network and a one-dimensional convolutional network respectively, and performing feature alignment and fusion; cross-modal dynamic coupling is realized through cross attention and a gating mechanism, and feature noise reduction is performed in combination with adaptive threshold filtering; feature weighted fusion is completed based on multi-level feature division and a channel attention mechanism, and a comprehensive time sequence feature vector is generated; and finally, through nonlinear feature enhancement, Softmax probability prediction, sliding window smoothing processing and majority voting, outputting a classification result aligned with a timestamp. The method has the advantages of being high in feature fine granularity, good in cross-modal fusion effect and the like, and is suitable for the fields of intelligent manufacturing, behavior recognition, medical monitoring and the like.

The invention relates to a security camera abnormal behavior identification method and system based on multi-modal fusion. The method comprises the steps of obtaining multi-modal data such as visible light image data, infrared image data and audio signal data of a security camera in a target monitoring period; performing feature extraction on the data of different modes to obtain respective feature vector sets; the features of all the modes are fused, and in the multi-mode feature fusion process, a weighted fusion algorithm for dynamically adjusting the fusion weight according to the confidence coefficient weight of feature vectors of all the modes is adopted; inputting the fusion feature vector set into an abnormal behavior classification model for classification and early warning; according to the scheme, the accuracy and effectiveness of abnormal behavior recognition in a complex environment can be improved, and then the overall efficiency of security monitoring is improved.

The invention discloses a single-person abnormal behavior identification method and system based on multi-modal skeleton feature fusion, and the method comprises the steps: S1, collecting continuous RGB images and infrared thermal imaging images in a monitoring video, carrying out the human body detection and key point estimation of visible light and infrared images through employing a multi-modal fusion model of YOLOv12 in combination with Transform, and constructing a single-person posture time series data set; s2, key point speed vectors are calculated for the continuous skeleton frame sequence of each target person, skeleton key point information and speed information are fused, and an action feature sequence is formed; s3, inputting the motion feature sequence into an MPED-RNN model, decomposing skeleton motion into a global displacement component and a local attitude deformation component, and performing joint coding, decoding and prediction through a dual-channel GRU network; and S4, calculating a prediction error and a reconstruction error according to a reconstruction result and a future skeleton key point prediction result, evaluating whether the current behavior deviates from a normal trajectory, and judging whether the current behavior is in an abnormal state. According to the invention, real-time identification of abnormal behaviors of a single person in a complex scene is realized.

The invention relates to the technical field of behavior recognition, and discloses a lightweight abnormal behavior recognition system and method based on edge computing, and the method comprises the steps: obtaining multi-dimensional heterogeneous behavior observation data covering a target region through an edge side behavior collection node system, and constructing a behavior dynamic feature vector set which can be iteratively updated; performing structure self-adaptive decoupling processing on the behavior dynamic feature vector set, and recording the fluctuation convergence rate of an abnormal category in real time in a model iteration process; judging the behavior recognition stability in the model reasoning stage, and constructing a behavior transfer trajectory map in combination with the environmental transaction interference factors; extracting density disturbance parameters of the abnormal behaviors, and generating a behavior intervention evaluation result set; and performing hierarchical risk judgment on the current identification behavior result, and automatically generating an edge execution regulation and control instruction set. The method has the advantage of improving the efficiency.

The invention discloses a poultry behavior abnormity real-time monitoring system based on multi-modal image fusion, particularly relates to the technical field of intelligent breeding behavior recognition, and is used for solving the problem of poor behavior monitoring accuracy under feather shielding. The method comprises the following steps: firstly, through combined perception of a visible light image and an infrared image, extracting a claw track interruption point and an anus temperature gradient direction, and realizing analysis of a motion state of a sheltered area; then, in combination with the heat conduction delay characteristic and the group movement direction, the flexion and extension angle of the covered leg joint is inverted, and a complete gait sequence is generated; thirdly, multi-source features such as gaits, temperature differences and body postures are fused, and a dynamic deviation model of the individuals relative to the mass center of the group is constructed; and finally, generating a stress behavior threshold curve according to the ground temperature and the ammonia gas concentration, outputting an abnormal behavior type and confidence, and realizing intelligent distinguishing of mechanical obstacles and adaptive behaviors.