62 results about "Visual flow" patented technology

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.

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.

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 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 invention provides a multi-mode document processing method, system and equipment, and the method comprises the steps: obtaining a multi-source heterogeneous file, and converting the multi-source heterogeneous file into a standard document; performing layout element analysis on the standard document, and determining a section set; according to each document semantic block in the section set, analyzing the spatial relationship, visual flow and semantic association between the document semantic blocks so as to generate a global content sequence conforming to a reading sequence; constructing an optimal analysis pipeline according to the type of each document semantic block in the global content sequence, analyzing each document semantic block based on the optimal analysis pipeline, and determining an analysis result; and generating the analysis result into structured data, and outputting the structured data. According to the method, the reading sequence is corrected, sequence logic coherence is ensured, and the semantic disjunction problem caused by only relying on the spatial position is solved; according to the method, scattered recognition results are associated and integrated and converted into structured data, so that the problem that information is scattered and difficult to reuse is solved.

Systems and methods are provided to power video conferencing and remote collaboration with subsymbolic and symbolic artificial intelligence. The autonomous video conferencing systems of this disclosure include one main smart camera and multiple peripheral smart cameras, optionally coupled with one or more smart sensors. Each smart camera is equipped with a vision pipeline supported by machine learning to detect objects and their interactions as well as related changes in gesture and posture, and a virtual director adapted to apply a predetermined rule set consistent with television studio production principles. The main camera is adapted to select and update a focus video stream in real time under the direction of its virtual director and stream the updated focus stream to a user computer. Methods for creating an automated television studio production for a variety of conferencing spaces and special-purpose scenarios with virtual director assistance are provided.

The application provides an IVR flow arrangement method, device and equipment and a storage medium. The method pre-establishes a node database, and a user can generate a node path from a first node to an end node of a flow by using the node database. The coordinates of each node on the node path are calculated by using a coordinate algorithm, the nodes containing the coordinate information are encapsulated, an interactive voice response (IVR) flowchart is generated and displayed, a complete set of IVR flows related to the first node is automatically generated based on the node database and the IVR flowchart, and the efficiency of IVR flow configuration is effectively improved. A visual flow customization environment is provided for the user, operations of the user on the IVR flowchart in the visual flow customization environment are received and processed, IVR flow arrangement data is obtained, and the efficiency and data accuracy of flow arrangement can be greatly improved.

Approaches are disclosed for enhancing frame rate and visual smoothness in real-time video streams through multi-frame interpolation. A classification neural network analyzes two sequential frames, outputting confidence scores that indicate the reliability of motion data for each pixel. These scores determine whether a pixel's motion is accurately described by motion vectors or should be treated as static. The classification results are reused to generate intermediate frames by warping the original frames based on the motion characteristics. Blending weights are calculated by combining warped motion vector confidence values with static values, and a second neural network refines the alignment and blending of candidate frames. This second network predicts intermediate flows and generates new blending weights, which are used to warp and blend the candidate frames, ultimately producing a final interpolated frame that enhances visual smoothness and consistency in the video stream.

The invention provides a time backflow type digital memory presentation method, and relates to the technical field of mobile internet, and the method comprises the steps: obtaining a starting time point and an ending time point which represent a to-be-presented memory time range; calculating to obtain a total time difference through the starting time point and the ending time point; based on the total time difference, setting a total step number required for realizing the reverse animation and an execution interval of each step, and obtaining an animation progress parameter according to the total step number and the execution interval; the animation progress parameter is initialized, so that the animation progress parameter is linearly changed from an initial value to a final value in an animation execution process; when each step of the animation is executed, obtaining a current value of an animation progress parameter as input; and applying the input current value to a preset slow motion function for calculation so as to obtain a nonlinear progress parameter with an S-shaped change curve. According to the invention, immersion, emotional resonance and visual fluency of memory presentation are improved.

Various implementations include methods for processing images from a device camera to identify potential malicious attacks on the camera. These implementations may include using multiple different trained image processing models or vision pipelines to process images received from the device's camera to obtain multiple different image processing outputs, and performing multiple consistency checks on the multiple different image processing outputs. Such consistency checks compare two or more selected outputs from the multiple different outputs to detect inconsistencies that may be associated with or caused by an attack on the camera. Indications of an attack on the camera may be reported to and considered by the device's autonomous driving system, or otherwise addressed in one or more mitigation actions.

The invention relates to the technical field of hydrological monitoring and computer vision, and discloses a visual flow measurement method, device and equipment based on a large-scale particle image flow measurement algorithm and a medium, and the method comprises the steps: obtaining the monitoring video data of a to-be-measured river and camera calibration parameters; image frames in the video data are preprocessed; performing pixel displacement calculation on the preprocessed continuous image frames through a large-scale particle image velocity measurement algorithm to generate a surface velocity vector field, and converting the surface velocity vector field into an actual velocity; performing multi-criterion chained filtering on the surface flow velocity vector field to obtain an optimized flow velocity field; mapping the optimized flow velocity field to a river section grid, and performing integral calculation based on section water depth data to obtain a section flow time sequence; and generating a transparent vector layer based on the optimized flow velocity field, fusing the transparent vector layer with the original video frame by frame, and outputting a flow field visualization video. According to the invention, efficient, accurate and stable non-contact river flow measurement is realized, and the processing efficiency and the result reliability are improved.

The application provides a kind of machine vision collaborative control method and system applied to intelligent processing, it is related to computer vision technical field.First, the association rule set of the visual feedback of processing equipment and control instruction is established, covering the mapping relationship of visual feature sequence and control instruction template and collaborative trigger condition, then based on the association rule set, the real-time collected processing area visual flow is carried out feature sequence extraction and matching processing, obtains control instruction template and collaborative trigger condition satisfaction parameter, then according to the above parameter generates dynamic control parameter adjustment sequence, the sequence is input into the collaborative control module of processing equipment, executes parameter loading and action calibration processing, finally, the mapping weight in association rule set is updated based on the processing result, and the collaborative control log is output, so that the real-time perception and accurate collaborative control of processing equipment in precision manufacturing are realized, and the quality and efficiency of part processing and docking are improved.

The application discloses a visual water flow speed measurement method in a non-uniform light field, comprising the following steps: S1, collecting video frames and converting the collected video frames into single-channel images; S2, performing mask operation based on the single-channel images, shielding irrelevant background interference and extracting ROI; S3, performing smooth filtering on the images after the mask operation and removing abnormal noise points; S4, performing local Gaussian threshold segmentation by using initial values to obtain binary images, wherein the initial parameters include neighborhood size and threshold bias; and S5, detecting feature points by using an ORB feature point detection algorithm on the binary images. Through the local Gaussian threshold segmentation technology, the threshold is dynamically calculated based on the local neighborhood gray scale characteristics of pixels instead of relying on a global fixed threshold, so that the non-uniform illumination space and time fluctuation caused by water surface reflection, building shadow, weather change and the like are effectively offset, the core assumption of the brightness constancy of the optical flow method is met, and tracking failure caused by the distortion of the imaging features of the target is avoided.

The invention relates to a visual process arrangement method and device, electronic equipment and a program product. The method comprises the steps that in response to visual arrangement of a plurality of graphic tools on a canvas, a graphic business process is obtained, and the graphic tools are used for representing nodes in the graphic business process; based on the business logic relationship of the graphical business process, determining structured data corresponding to each node in the graphical business process; and on the basis of the structured data, an EL expression used for realizing a graphical business process function is generated, and the EL expression is used for representing the business logic relation between the nodes. According to the method and the device, the graphical tool is visually arranged, the EL expression is dynamically generated according to user requirements, and the function of the graphical business process is realized through the EL expression.

The invention discloses a multi-mode and multi-agent oriented visual arrangement system. The system comprises a visual arrangement module, a visual process compiling module, a multi-mode data bus module, a self-adaptive intelligent arrangement engine and a visual presentation module. According to the invention, the core pain point in processing complex multi-mode and multi-agent tasks in the prior art is effectively solved, and a powerful, efficient and easy-to-use development and operation platform system is provided for constructing next-generation enterprise-level intelligent application.

The utility model discloses an air cooling device of a cylindrical lithium battery, which comprises a first visual flow channel, a second visual flow channel, an air rectifying plate, a battery pack and an airflow isolation box body, and the flow channels and the box body are made of transparent materials, so that the airflow distribution and the cooling effect can be conveniently monitored; the air rectifying plate is provided with a plurality of honeycomb strip-shaped hollow structures, so that the air flow path is optimized, and the cooling efficiency is improved; the battery pack device is modularized in structure, has good adaptability and expansibility, and is suitable for distribution of lithium battery packs of different specifications. The device has an effective and efficient heat dissipation function and is suitable for the fields of electric automobiles, energy storage equipment and the like.

The invention relates to the technical field of electric energy storage, in particular to a dynamic network flow monitoring visualization method and system based on a network construction SVG, and the method comprises the steps: collecting power flow monitoring data through a step S1, carrying out the data processing of the power flow monitoring data through a step S2, and predicting the flow trend through a step S3. The method comprises the steps of obtaining static data, obtaining a visual flow SVG graph through a step S4, converting the static data into a real-time response and interaction visual power flow SVG graph through a step S5, carrying out quantitative evaluation on an abnormal degree through a step S6, combining with a multi-level alarm rule, and carrying out alarm processing through a step S7, so as to improve the reliability and the monitoring efficiency of power network flow monitoring.

This invention discloses an AI vision-based drainage pipe network siltation risk analysis system, belonging to the field of intelligent monitoring technology for drainage pipe networks. It acquires continuous sequence images of the drainage pipe interior through a visual acquisition module and preprocesses them using advanced image processing technology to ensure image quality. The AI ​​visual feature extraction module employs a lightweight convolutional neural network to accurately quantify and extract three core visual features: pipe wall deposits, siltation contours, and water flow interfaces. An implicit coupling modeling module implicitly couples visual features with unsteady flow dynamics equations, establishing a mapping relationship between visual features and fluid parameters. A pure visual flow regime inversion module uses only visual feature parameters as input to achieve pure visual flow regime deduction without relying on external flow sensors, inverting local flow resistance and velocity distribution. Finally, a siltation risk assessment module constructs a quantitative model based on the flow regime inversion results, completing the quantitative assessment, level classification, and early warning output of siltation risk.

The present disclosure belongs to the technical field of nuclear power, and particularly relates to a nuclear power digital maintenance procedure structuring method and device. The present disclosure converts the natural language description of the maintenance procedure into a structured procedure flow chart through a visual flow chart editing mode, dynamically associates each step with related data in a multi-type data storage system, forms a seamless set of multi-element data, and finally generates a digital procedure object that can be parsed and executed by a computer, which is friendly in interface and convenient in operation, and lays a foundation for subsequent intelligent application.

The application relates to a pig posture classification method and system based on a double-flow network. First, pig image is collected, and key point information of the pig is extracted from the pig image. Then, based on the key point information, a posture stream network is used to generate posture structure features. Meanwhile, a visual stream network is used to extract global visual features from the pig image. Then, the posture structure features and the global visual features are fused to obtain fused features. Finally, the posture of the pig is classified based on the fused features. The application uses key point extraction and graph neural network to extract pig posture structure features, effectively captures the topological relationship between joints, processes the posture classification result in time sequence consistency, uses a sliding window filtering algorithm to correct the instantaneous misclassification caused by video noise and other factors, uses the time sequence information of adjacent frames to infer the posture of the pig in a severely occluded or lost frame, and significantly improves the accuracy of pig posture classification.

A method and apparatus of processing a video data stream, a method and apparatus of processing an image, and a storage medium are provided, which may be applied to artificial intelligence technology, and in particular to computer vision, streaming media, etc. The method includes: determining N video data streams from N image acquisition devices, where each video data stream includes at least one data package, and N is an integer greater than or equal to 2; dividing a plurality of data packages in the N video data streams into N data package sets, where each data package set corresponds to a video data stream; determining N first queues respectively corresponding to the N data package sets, where each first queue contains a first storage address in a first storage space for each data package in a data package set corresponding to the first queue; and sending the N first queues.

The present application provides a kind of based on multi-source vision sensor fusion's blast furnace slag iron flow detection method and system, belongs to blast furnace slag iron flow detection technical field, including: hyperspectral image data input slag iron ratio estimation model, the composition of molten iron and slag is identified and separated, extract the spatial distribution information of iron phase and slag phase, calculate slag iron ratio;Slag iron flow time series data input time series visual flow velocity perception model, the dynamic analysis of slag iron flow flow process, extract time evolution characteristics and flow trajectory information, calculate slag iron flow velocity;Three-dimensional point cloud data input three-dimensional point cloud reconstruction flow section analysis model, the spatial form of slag iron flow is modeled and analyzed, extract three-dimensional geometric structure and cross-sectional shape characteristics, calculate slag iron flow cross-sectional area.Slag iron ratio, slag iron flow velocity, slag iron flow cross-sectional area are jointly modeled, the adaptive calculation of slag iron flow mass flow is carried out, and slag iron flow detection result is output.The present application can effectively carry out slag iron flow detection.

The application provides an end-to-end automatic driving trajectory generation method, system, computer device and medium, belongs to the field of path intelligent planning, and comprises the following steps: acquiring traffic environment three-dimensional perception information and a natural language instruction; converting the traffic environment three-dimensional perception information to obtain a structured semantic label, and adjusting the weight of the structured semantic label through the natural language instruction; extracting a visual flow feature from the structured semantic label based on the weight of the structured semantic label, and extracting a language flow feature from the natural language instruction; fusing the visual flow feature and the language flow feature based on a cross-modal attention mechanism to obtain a fusion feature and generate a simulation trajectory; calculating a real-time collision time, and when the real-time collision time is greater than a collision time threshold, modifying the simulation trajectory based on a physical constraint and a traffic regulation library to obtain a modified driving trajectory. The automatic driving simulation trajectory constructed by the method conforms to the user's intention, and the system reliability and human-computer cooperation efficiency are improved.

The invention discloses a refueling robot cooperative control system based on a limited guide rail and a dynamic operation planning method, and belongs to the technical field of robot automatic control. The system comprises a U-shaped guide rail, a movable chassis, a multi-degree-of-freedom mechanical arm, a visual perception device and a central controller. According to the method, the position and posture of the oil port are dynamically calculated through visual flow in the advancing process of the movable chassis, the optimal parking point is preferably selected from discrete mechanical limiting points on a guide rail on the basis of inverse kinematics simulation, and the problem that traditional fixed limiting cannot adapt to vehicle parking errors is solved; meanwhile, an anisotropic variable-stiffness impedance control strategy is adopted in the oil gun butt joint stage, and smooth insertion is achieved by reducing radial stiffness. Cooperation of perception and flexible operation during advancing is achieved, the low-cost advantage of the guide rail scheme is kept, meanwhile, the environment adaptability and operation efficiency of the system are remarkably improved, and the method is suitable for automatic transformation of gas stations.

Approaches to improving frame rate and visual smoothness in real-time video streams through multi-frame interpolation are revealed. A neural classification network analyzes two consecutive images and outputs confidence scores indicating the reliability of motion data for each pixel. These scores determine whether the motion of a pixel is accurately described by the motion vectors or should be treated as static. The classification results are reused to generate intermediate frames by warping the original images based on their motion properties. Blended weights are calculated by combining the confidence scores of the warped motion vector with static values, and a second neural network refines the alignment and blending of the candidate images.This second network predicts intermediate streams and generates new blend weights that are used to warp and blend the candidate images, ultimately producing a final interpolated image that improves the visual smoothness and consistency in the video stream.

This invention relates to the field of visual analytics technology, and discloses a method and apparatus for flow field visual analysis driven by a large language model. The method includes: acquiring a flow field dataset to be analyzed and user commands; performing semantic analysis on the user commands using a large language model to obtain structured commands; based on the structured commands, retrieving corresponding target data from the flow field dataset using the large language model, and generating descriptive text corresponding to the target data; running a visualization agent based on the target data and the descriptive text to obtain visualization results; and displaying the target data according to the data display method indicated by the user commands. By using a large language model to achieve natural language-based interaction, user operation commands or text commands are converted into structured commands, thereby presenting visualization results and explanations in real time according to user needs, greatly improving the user experience.

The invention discloses an animation frame interpolation method and system based on image data, and relates to the technical field of data processing. Touch interaction parameters are collected, a key interface frame is obtained when the estimated pixel displacement is larger than a visual fluency displacement threshold value, initial motion optical flow features are mapped, an interaction control active area is extracted, and the interaction control active area is extracted; extracting a local structure feature of the key interface frame, extracting a global layout context dependent feature under the guidance of the active area of the interactive control, and fusing the local structure feature and the global layout context dependent feature to generate an interface context content feature; and inputting a multi-mode splicing feature formed by splicing the initial motion optical flow feature and the interface context content feature into a joint selective state space model, calculating a correlation weight, fusing a joint space fusion feature, predicting an optical flow residual error to generate a refined target motion optical flow, and twisting a key interface frame based on the refined target motion optical flow to obtain a target motion optical flow. And calculating complexity to synthesize an intermediate transition frame, and outputting the intermediate transition frame to a display panel for rendering.

This invention discloses a naked-eye 3D augmented reality interactive display system, relating to the field of naked-eye 3D display and real-world interaction technology. The invention acquires and fuses user limb depth images through a multi-source depth data acquisition module, completes the mapping of depth and parallax and viewpoint correction through a parallax physical depth mapping module, divides regions to determine occlusion types through an occlusion relationship determination module, generates a dynamically smooth occlusion layer through an adaptive occlusion layer generation module, and displays the occlusion effect fusion display module by superimposing the occlusion layer with multi-viewpoint virtual images. This effectively solves the problems of inaccurate occlusion relationships between user limbs and virtual objects and visual abruptness in naked-eye 3D interaction, significantly improving the spatial consistency and visual smoothness of the interaction, and achieving a realistic and natural naked-eye 3D augmented reality interactive experience.

The invention discloses a RAW image processing method and device, computer equipment and a storage medium, and the method comprises the steps: loading a to-be-processed RAW image, and carrying out the initialization of the RAW image; responding to an operation instruction of a user, and obtaining an operation type corresponding to the operation instruction; and performing corresponding processing on the RAW image based on the operation type, and performing real-time rendering on the RAW image in combination with a pyramid hierarchical frame insertion rendering mode and an asynchronous scheduling strategy. According to the embodiment of the invention, the exchange law characteristic of the algorithm is utilized, the calculation sequence is dynamically adjusted, and'one-time noise reduction and multiple-time color modulation 'is realized, so that the calculation load in a high-frequency interaction scene can be greatly reduced. Meanwhile, through a pyramid grading and task preemption mechanism, the visual fluency during user operation is ensured, low-resolution preview is preferentially displayed, and smooth transition to a high-definition image is realized by utilizing the visual persistence characteristic of human eyes, so that the problems of jamming and resource waste during real-time color matching in the prior art are solved.

The invention belongs to the technical field of mining machinery automation control, and particularly relates to an underground high-precision robot control method and system based on machine vision, and the method comprises the steps: tracking feature points through a sparse optical flow algorithm, calculating the visual flow field consistency confidence coefficient through the discrete degree of the motion direction angle of the feature points, and calculating the visual flow field consistency confidence coefficient; the interference of underground dust is filtered; combining the visual flow field consistency confidence coefficient with a phase plane constructed by a position error and a speed error, and calculating a jitter energy divergence index to evaluate the nonlinear jitter intensity of the mechanical arm; and based on the jitter energy divergence index, dynamically calculating a real-time damping coefficient through a nonlinear adaptive adjustment law, and adjusting the damping parameter of the controller in real time. According to the method, environmental noise is filtered through physical optical flow field statistical characteristics, damping characteristics are dynamically adapted, and vibration resistance, fast response and stable hovering of the mechanical arm under complex working conditions are achieved.