154 results about "Visual flow" patented technology

In some embodiments, a multimodal computing system receives a query and identifies, from source documents, text passages and images that are relevant to the query. The multimodal computing system accesses a multimodal question-answering model that includes a textual stream of language models and a visual stream of language models. Each of the textual stream and the visual stream contains a set of transformer-based models and each transformer-based model includes a cross-attention layer using data generated by both the textual stream and visual stream of language models as an input. The multimodal computing system identifies text relevant to the query by applying the textual stream to the text passages and computes, using the visual stream, relevance scores of the images to the query, respectively. The multimodal computing system further generates a response to the query by including the text and / or an image according to the relevance scores.

The invention provides an end-to-end automatic driving track generation method and system, computer equipment and a medium, and belongs to the field of path intelligent planning, and the method comprises the steps: obtaining traffic environment three-dimensional perception information and a natural language instruction; converting the traffic environment three-dimensional perception information to obtain a structured semantic mark, and adjusting the weight of the structured semantic mark through a natural language instruction; based on the weight of the structured semantic mark, visual flow features are extracted from the structured semantic mark, and language flow features are extracted from the natural language instruction; based on a cross-modal attention mechanism, the visual flow features and the language flow features are fused, fusion features are obtained, and a simulation track is generated; real-time collision time is calculated, when the real-time collision time is larger than a collision time threshold value, the simulation track is corrected based on physical constraints and a traffic regulation library, and a corrected driving track is obtained. The automatic driving simulation track constructed by the method conforms to the intention of a user, and the system reliability and the man-machine cooperation efficiency are improved.

The invention relates to an unmanned aerial vehicle self-adaptive wind-resistant flight control method and system based on multi-mode environment perception. The method comprises the steps that a miniature wind speed and direction sensor array obtains real-time wind speed and wind direction information of the surrounding environment of an unmanned aerial vehicle; the inertial measurement unit obtains real-time attitude angle, angular velocity and acceleration information of the unmanned aerial vehicle; the visual flow sensor obtains real-time image data of the surrounding environment of the unmanned aerial vehicle; the data fusion and wind field estimation module receives multi-source data collected by the miniature wind speed and direction sensor array, the inertial measurement unit and the visual flow sensor, carries out fusion processing on the multi-source data based on an extended Kalman filtering algorithm, and predicts to obtain predicted wind field information in a future short time window; the flight control core module adopts an adaptive sliding mode control algorithm to calculate a compensation control quantity for counteracting wind disturbance, and performs real-time control on the unmanned aerial vehicle based on the compensation control quantity; and the wind resistance and the operation reliability of the unmanned aerial vehicle in a complex dynamic environment are improved.

The invention relates to the technical field of artificial intelligence, in particular to an AI digital human-based customer service method and system, and the method comprises the steps: receiving a voice stream, a visual stream and auxiliary sensor data inputted by a user, achieving the time-space alignment of multi-modal data through a cross-modal feature synchronization engine, and fusing the multi-modal data into a unified fusion data packet; based on the fused data packet, extracting an emotion intensity value and an emotion type code in real time, generating a dynamic emotion label, and extracting service demand characteristics to generate a demand vector; inputting the dynamic emotion label and the demand vector into a strategy decision tree, and generating a strategy identifier containing an emotion pacifying coefficient and a response emergency degree; the AI digital human is driven to generate the multi-modal response based on the strategy identifier, deep perception and flexible response to the user state can be achieved, the interaction naturalness and the user satisfaction of a digital human service system are effectively improved, and the method is suitable for multiple intelligent service scenes such as online customer service and intelligent terminal assistants.

The invention provides a service robot cross-modal user intention analysis method and system. The method comprises the following steps: firstly, acquiring and analyzing a voice instruction sequence and a visual flow containing a gesture, fusing a task target object and a gesture pointing direction to generate an intention hypothesis, and then projecting the intention hypothesis to a knowledge graph of an environment with a body; the method comprises the following steps: acquiring an intent hypothesis of a user, querying a candidate object set simultaneously matched with spatial constraints and semantic constraints of the intent hypothesis, when ambiguity exists in the candidate object set, generating an inquiry instruction for eliminating the ambiguity according to a relation path in a knowledge graph of the body environment, and finally analyzing and confirming a final user intent according to the response of the user to the inquiry instruction; according to the technical scheme provided by the invention, the ability of the robot to cope with uncertainties such as unknown reference and environment shielding in a real scene is enhanced, the interactive experience closer to the human cognitive level is realized, and the interactive efficiency and satisfaction of the user in a complex environment are also improved.

The invention relates to the technical field of flow batteries, in particular to a flow battery flow field uniformity detection device and a detection method, and the flow battery flow field uniformity detection device comprises a visual flow field assembly, a constant temperature water tank, a circulating pump, a pressure gauge, a thermal imager and a connecting pipeline. The visual flow field assembly, the constant-temperature water tank, the circulating pump and the pressure gauge are connected through connecting pipelines, the shooting range of the thermal imager covers the whole visual flow field area, and the temperature change of the flow field area along with time is recorded. According to the invention, the temperature change rate of the fluid at the outlet is controlled by adjusting the heating power of the water tank, and when the fluid passes through the flow channel of the electrode frame, a relatively large temperature gradient is ensured in the flow channel range of the whole visual flow field assembly. And calculating a temperature deviation value to evaluate the fluid uniformity according to the linear relationship between the temperature difference and the speed difference.

The invention relates to the field of image processing, in particular to an industrial feeding anomaly recognition method based on visual flow detection, and the method comprises the steps: obtaining an image sequence for optical flow calculation through the collection and preprocessing of an image sequence of an industrial feeding port; performing nonlinear fusion on the pixel brightness value and the time brightness change rate to obtain a highlight suppression factor; carrying out distribution analysis on the local gradient structure tensor characteristic value to obtain a gradient structure confidence factor; performing joint weighting on the highlight suppression factor and the gradient structure confidence factor to obtain a dynamic space-time confidence weight; weighted optical flow calculation and variance threshold judgment are carried out on the dynamic space-time confidence coefficient weight, a feeding abnormal state recognition result is obtained, and therefore the problems of optical flow calculation distortion and abnormal misinformation caused by mirror surface highlight in the SMT micro component feeding process are solved.

The invention relates to the technical field of computers, and provides a flow chart processing method and device, equipment and a storage medium, and the method comprises the steps: obtaining a to-be-processed visual flow chart; flow nodes of the visual flow chart represent logic units, and connecting lines between the flow nodes represent an execution sequence and a dependency relationship between the logic units; extracting flow information of the visual flow chart; the flow information comprises node types and node attributes of the logic units and a logic relation between nodes; and converting the visual flow chart into executable codes or scripts based on the flow information. By analyzing the flow chart, the flow information of the flow chart is directly converted into the executable code or script, so that the logic arrangement of the flow chart is simplified, the code development difficulty of the logic arrangement of the flow chart is reduced, and the logic arrangement efficiency of the flow chart is improved.

The invention discloses a finance and taxation field agent workflow arrangement method and system based on a natural language, and the method comprises the steps: S1, widely collecting workflow arrangement cases of a finance and taxation field typical business scene, carrying out the detailed analysis of each case, and extracting a key business process and node information, S2, taking all structured workflow data as nodes, and carrying out the calculation of the workflow arrangement cases, constructing a finance and tax workflow knowledge graph by taking the semantic similarity, the function multiplexing relationship and the business process dependency relationship among all the structured workflow data as edges; s3, based on structured data and natural language description corpora constructed by the finance and tax workflow knowledge graph, performing training optimization on the base large language model by adopting a domain self-adaption and task specific fine tuning technology; s4, a user inputs a natural language demand through the interactive interface, and the natural language demand is processed by using the demand analysis agent, the retrieval agent and the arrangement agent; and S5, displaying the generated draft workflow in a visual flow chart form.

The invention provides a soft soil stratum slurry balance shield slurry feeding / discharging pipeline slurry flow test system. The system comprises a slurry circulation pipeline, monitoring equipment, a soil body cutting simulation system, a PIV-based visualization device, an analysis system and a control system. The slurry circulating pipeline comprises a slurry tank, a vortex pump and a slurry conveying pipe; the monitoring equipment comprises a plurality of sensors; the soil body cutting simulation system is used for simulating a shield excavation working condition; the visual device based on PIV is used for recording a visual flow field of slurry flowing in the slurry conveying pipe in real time; the analysis system is used for macroscopic and microscopic analysis; and the control system controls equipment power in the system. The device and the method can simulate the working conditions of most slurry balance shield slurry discharge pipelines and record macro and micro evaluation indexes in the slurry flowing process, so that the flowing property and the slag carrying capacity of slurry in the slurry feeding / discharging pipelines under different stratum conditions are explored, and the device and the method are low in cost, high in operability and high in compatibility.

The invention provides a fire prevention and control method and system using computer vision, and relates to the technical field of computer vision, and the method comprises the steps: firstly obtaining a monitoring scene visual flow containing visible light and infrared thermal imaging images, and carrying out the visual flow segment correlation modeling of the monitoring scene visual flow to obtain a correlation chain cluster; a fire evolution feature set is obtained through feature linkage evolution analysis, a fire prevention and control response instruction containing prevention and control operation guidance information is generated based on the fire evolution feature set, the fire prevention and control response instruction is sent to an execution terminal, feedback is received, and comprehensive monitoring and effective prevention and control of the fire are achieved. Fire prevention and control efficiency and reliability are improved, and fire loss is reduced.

The invention discloses a heat transfer process simulation method and device in visual flow and a storage medium, and the method comprises the steps: (1) determining a full-field source item f (x) and a thermal function Hb (x) value on a boundary according to the data of a flow velocity field and a temperature field; (2) carrying out collision on the distribution function ni (x, t) under the action of a full-field source item f (x) to obtain # imgabs0 #; (3) processing distribution function collision at a boundary; (4) migrating the distribution function # imgabs1 # after collision to the adjacent grids according to the discrete grid speed; (5) calculating a thermal function H (x, t) = sigma ini (x, t) according to the distribution function; (6) calculating the change of a full-field thermal function, and carrying out termination judgment; and (7) visualizing the hot line distribution according to the thermal function distribution. The method has good parallelism performance and a simple boundary format, and meanwhile has good applicability to boundary non-body-fitted grids.

The invention relates to the technical field of computers and simulation analysis, in particular to a flow field simulation analysis method in the starting process of a water turbine, which comprises the following steps of: performing geometric pretreatment and grid pretreatment on a movable guide vane of the water turbine, and constructing a simulation environment; constructing a basic flow simulation model of the movable guide vane of the water turbine, adding a turbulence equation to model a turbulence phenomenon in the operation process of the water turbine, and inputting the turbulence phenomenon into the simulation environment for flow field simulation; a water turbine and a steady-state solver are initialized, and transient simulation results of flow velocity and pressure in the starting process of the water turbine in the simulation environment are solved in combination with analysis of the starting transition process of the water turbine until calculation convergence is achieved; and post-processing the calculation result, and outputting a visual flow field result. The variable speed transient simulation process of the water turbine starting process can be achieved, and high-precision transient numerical simulation of flow field changes in the water turbine starting process is achieved.

The invention discloses a naked-eye 3D augmented reality interactive display system, and relates to the technical field of naked-eye 3D display and reality interaction. According to the invention, the multi-source depth data acquisition module acquires and fuses a user limb depth image, the parallax physical depth mapping module completes depth and parallax mapping and visual angle correction, the occlusion relation determination module divides regions to determine the occlusion type, and the adaptive mask layer generation module generates a dynamic smooth mask layer. The shielding effect fusion display module is used for overlapping and synthesizing the shielding layer and the multi-view virtual image and then displaying the overlapped and synthesized shielding layer and the multi-view virtual image; the problems that in naked eye 3D interaction, the occlusion relation between the limbs of the user and the virtual object is not accurate, and the vision is abrupt are effectively solved, the space consistency and the vision fluency of interaction are remarkably improved, and the real and natural naked eye 3D augmented reality interaction experience is achieved.

The invention relates to the technical field of computers and the field of financial science and technology businesses, and discloses an application code visualization method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining source codes corresponding to all source files, and generating a code visualization plug-in according to each source code; and generating a visual flow chart by analyzing the logic flow and the target transfer function of the code visualization plug-in, and displaying the visual flow chart. By means of the mode, the source code is read through the code visualization plug-in and converted into the representation form of the visualization flow chart, the visualization flow chart assists a developer to understand the code logic and structure more quickly, and when codes are modified or reconstructed, the visualization flow chart provides a clear reference. The code visualization efficiency of the front-end application in the financial science and technology business field is improved, the visualization flow chart enables a user to quickly understand an existing code structure, and development of new functions and improvement of existing functions are accelerated.

The invention belongs to the technical field of relay protection practical training and evaluation of a power system, and particularly relates to a relay protection operation coach and evaluation system and method based on AR glasses, and the system comprises an AR intelligent glasses terminal which is worn by an operator and is used for collecting the depth visual flow data of a field environment in real time; an edge computing server which is in communication connection with the AR intelligent glasses terminal and is used for obtaining the visual flow information collected by the AR intelligent glasses terminal, constructing a state sensing model based on a power equipment topological relation and realizing direct semantic level identification of screen cabinet elements; meanwhile, generating a three-dimensional visual operation indicator, a dynamic safety early warning signal and real-time evaluation feedback information, and superposing the three-dimensional visual operation indicator, the dynamic safety early warning signal and the real-time evaluation feedback information in the visual field of the AR intelligent glasses terminal; the cloud management platform is in communication connection with the edge computing server and is used for performing user management, storage and updating of an operation instruction knowledge graph, storage and analysis of historical evaluation data and unified allocation of system resources; the defects of a traditional training and evaluation system are overcome.

The invention relates to the technical field of multi-scene dynamic orchestration, in particular to a multi-scene-based orchestrable unified preprocessing rule method and system, and the method comprises the steps: carrying out the dynamic matching of a rule according to scene information, employing a visual rule orchestration engine, and employing a micro-service architecture design; the method has the beneficial effects that different service scenes can be covered through the modular rule base and the dynamic configuration capability, repeated development of a traditional single scene preprocessing system is avoided, and the technology reuse rate is improved. Visual flow arrangement is adopted, a user is allowed to customize a rule execution sequence, conditional branches and combinatorial logic, complex business demand changes are flexibly dealt with, and the technology coupling degree is reduced.

The present disclosure provides a video data stream processing method and device, an image processing method and device, and a storage medium, which can be applied to the technical field of artificial intelligence, and in particular to the technical fields of computer vision, streaming media, etc. The video data stream processing method comprises: determining N video data streams from N image acquisition apparatuses, each video data stream comprising at least one data packet, and N being an integer greater than or equal to 2; dividing a plurality of data packets included in the N video data streams into N data packet sets, each data packet set corresponding to one video data stream; determining N first queues respectively corresponding to the N data packet sets, each first queue comprising a first storage address of each data packet in the data packet set corresponding to the first queue in a first storage space; and sending the N first queues.

An AI-enabled visual workflow and model is provided for supporting interactions across multiple learning environments and systems. An admin can access a visual workflow editor to create workflows using the visual flow-based model. These workflows can be stored as workflow templates consisting of textual representations of interconnected nodes and any inputs, parameters, and outputs for the nodes. When a workflow is assigned to a student, the corresponding workflow template can be converted into an object model that a workflow processor can execute to dynamically generate a user interface representing the workflow. Workflows can be configured to include metaverse nodes by which a metaverse can be launched to enable the students to play as part of completing the workflows. Achievement nodes in the workflows can be configured to cause the corresponding achievements to be made available as virtual achievements in the metaverse.

The invention relates to the technical field of intelligent task management, in particular to an intelligent task management method and system based on visual process orchestration, and the method comprises the following steps: S1, task definition and decomposition: defining basic attributes and requirements of a task in the system, and decomposing a complex task into a plurality of subtasks; and S2, flow arrangement and visualization: arranging task flows by utilizing a visualization tool, setting an execution sequence, a dependency relationship and parallel or serial execution of the tasks. According to the intelligent task management method and system based on visual flow arrangement, the problems that in the prior art, an intelligent task management system cannot perform visual flow arrangement in the actual use process and lacks an intuitive visual interface, so that information such as task progress, resource allocation and flow state is difficult to be clear at a glance, and the task management efficiency is improved are solved. A large amount of time needs to be spent on looking up documents or carrying out verbal communication so as to know the latest dynamic problem of the task.

The invention discloses a space-time image texture angle detection and discrimination method and system, and belongs to the field of visual flow measurement instruments. According to the design, a Zynq series processor with an ARM + FPGA architecture is used as a core, and a system module comprises a PS host control end, a PL end texture enhancement module, a two-dimensional Fourier transform module, a frequency spectrum logarithm transform module, a Gaussian filter module and an image segmentation and texture angle detection and judgment module. Wherein the texture enhancement module inhibits noise interference and enhances the continuity of effective textures; the two-dimensional Fourier transform module converts the information of the spatial domain image to a frequency domain; the frequency spectrum logarithmic transformation module is used for carrying out logarithmic magnitude spectrum calculation on the transformed spectrogram so as to improve the visibility of a low-amplitude component; and the image segmentation and texture angle detection and discrimination module carries out angle detection and discrimination after effective and invalid signals are separated, and finally outputs an angle detection result. The method can be applied to a river surface flow velocity measurement system based on a space-time image velocity measurement method, and the adaptability and efficiency of flow measurement in complex scenes can be improved.

The invention is suitable for the technical field of hydrographic survey, and provides a millimeter wave radar speed measurement calibration method based on visual flow measurement, which comprises the following steps: acquiring data, constructing a database and preprocessing an image, performing target detection on the preprocessed image based on YOLO v8, performing region segmentation on a water surface region after target extraction is completed, and obtaining a millimeter wave radar speed measurement calibration result. Visual flow measurement calculation is carried out based on an optical flow method, when optical flow calculation is carried out, calculation should be carried out in each area, space-time synchronization is carried out on a visual flow measurement result and a flow velocity measured by a millimeter wave radar, credible areas are screened, abnormal data are processed, and linear coefficients are fitted for each credible area k through a least square method; according to the method, under the condition that millimeter wave radar speed measurement single-point measurement is limited, area speed measurement information is provided through visual flow measurement, and the problem that the millimeter wave measurement range is insufficient is solved.

The invention relates to a process modification auxiliary method based on a workflow engine. The process modification auxiliary method based on the workflow engine can greatly reduce the risk of process modification and improve the robustness of a service system. Besides, the influence range of process modification can be reflected simply and clearly through the visual flow chart, so that the user can focus on the affected area, and the safety of business circulation is improved. The method has the specific effects that the efficiency and accuracy of process modification are improved, and manual intervention and errors are reduced; the influence range of flow modification is accurately determined through the flow chart, and the modification risk is reduced; automatic modification and simulation test functions ensure stability and reliability after process modification.

The invention relates to the field of computer vision and artificial intelligence, in particular to a multi-modal fusion document image classification method based on multi-branch deep convolution and hierarchical semantic modeling. According to the method, visual and text information of a document is co-processed through a double-flow architecture; a visual semantic feature of a document image is efficiently extracted from a visual flow by adopting a lightweight convolution operation and multi-scale feature fusion mechanism; an original text is obtained from a text stream through an optical character recognition technology, text error correction and word embedding extraction are carried out through a fine tuning language model, word-level and sentence-level context modeling is realized by inputting a hierarchical semantic coding network, and document-level features are generated. And high-precision document image classification is realized by fusing the double-flow feature vectors and combining a full-connection layer. Experimental results show that the method has an excellent multi-modal feature fusion effect, relatively high classification precision and strong robustness to complex document scenes in a document image classification task, and is suitable for application scenes of various document image classification.

This invention relates to the field of construction safety monitoring and intelligent visual analysis technology, specifically a wiring-free construction safety monitoring system based on visual algorithm analysis. The system includes: an independent power supply module, a visual perception module, a scene risk topology assessment module, a computing-risk coupling scheduling module, a multi-level wake-up execution module, and a context-related early warning output module. It collects initial visual flow data, remaining power, and predicted charging power of the target monitoring area; divides the topology regions into different risk levels through semantic segmentation and assigns risk weights; dynamically generates computing power scheduling instructions containing the target sampling frame rate and target model complexity; outputs the target violation event identification results; and generates the final early warning result by combining the spatial inclusion relationship between the occurrence location and the topology region boundary. This achieves priority protection for high-risk spatiotemporal nodes and reduces ineffective energy consumption and invalid alarms in low-risk scenarios.

Generating and running computer vision pipelines for processing images and / or video. A user may select blocks for the computer vision pipeline, including image transformation blocks that transform images and / or video, and prediction blocks that detect objects or classes of objects in images and / or video. The user may specify an input source for the images and / or video and an output destination for the computer vision pipeline. Upon activation, the computer vision pipeline receives input images and / or video. The input images and / or video are transformed using the image transformation block to produce transformed images and / or video. Objects or classes of objects in the transformed images and / or video are detected using the prediction block to produce detected objects or classes of objects. The detected objects or classes of objects are transmitted to the output destination.

The invention aims to provide a gas turbine blade unstable flow visualization testing method, and belongs to the technical field of gas turbine experimental testing. Comprising the following steps: inputting boundary conditions and geometric conditions of a calculation model; numerical simulation flow field data under the condition are obtained through CFD calculation; if yes, the next step is carried out, and if not, a higher-quality grid is generated and returned to the encrypted grid, and CFD numerical simulation calculation is carried out again; selecting a proper resolution ratio based on an actual demand; generating a data source quantity and obtaining CFD calculation data according to the data source quantity; and outputting the three-dimensional transient visual flow field at the moment obtained by the checked high-precision CFD numerical simulation. The actual working condition of the gas turbine blade can be simulated, the test error of an actual instrument under the severe working condition is avoided, and the test difficulty is reduced. The precision of CFD numerical simulation is optimized, and the error is reduced, so that the credibility of the three-dimensional visualization cloud picture is improved.

The utility model discloses a visual flow adjusting device for feeding nitrogen during glue making, which comprises a glue making tank gas inlet pipe, a transparent pipe outer supporting pipe, a lower connecting flange pipe, a gas inlet valve and a notch groove formed in one side of the transparent pipe outer supporting pipe which are sequentially and hermetically connected from top to bottom, the transparent inner pipe is embedded in the transparent pipe outer supporting pipe and arranged at the position where the notch groove is located, the upper net plate is arranged between the glue making tank air inlet pipe and the transparent pipe outer supporting pipe, the lower net plate is arranged between the transparent pipe outer supporting pipe and the lower connecting flange pipe, and the flow ball is arranged in the transparent inner pipe and arranged between the upper net plate and the lower net plate. A plurality of mesh holes are formed in the upper mesh plate and the lower mesh plate; the air inlet valve is externally connected with nitrogen; the transparent inner tube is connected with the transparent tube outer supporting tube in a sealed mode. According to the scheme, the device has the advantages of simple structure, intuition, reliability and the like, and has very high practical value and popularization value in the technical field of glue making.

The invention discloses a logging instrument drilling fluid outlet flow monitoring system and method based on machine vision, and belongs to the technical field of drilling fluid detection. Collecting data; carrying out image convolution processing; extracting features; calculating a visual modal visibility probability; calculating a flow modal probability; calculating a pressure modal probability; determining a comprehensive cutoff probability based on multi-modal probability fusion; and determining a risk level and a hierarchical response mechanism based on the comprehensive cutoff probability. According to the method, three heterogeneous modes of vision, flow and pressure are fused, so that the problem that the reliability in the prior art is still difficult to meet the real-time decision-making requirement under the complex working condition is solved, the system reliability under the severe working condition is improved, and the technical effect of reducing the false alarm rate and the missing report rate is achieved.

The embodiment of the invention provides an electric power high-altitude operation peccancy detection method, and relates to the technical field of electric power system safety monitoring. The detection system comprises the following steps: acquiring a visual flow, and acquiring a to-be-detected image according to the visual flow; obtaining a backlight image in the image, and extracting a dark area in the backlight image; performing multi-scale transformation on the obtained dark area to enhance the dynamic range of the dark area; the converted dark area is input into a Light Net model, and a corresponding illumination image is generated; combining a dark area with the illumination image to generate enhancement; taking the non-backlight image and the enhanced image as a first image; judging whether the first image is shielded or not; and under the condition that the first image is not shielded, performing peccancy detection on the first image, and giving an alarm when a peccancy condition is detected. According to the invention, through backlight image enhancement, cross-modal data fusion detection and shielding compensation, the detection precision of the safety protection equipment under a complex working condition is effectively improved.