220 results about "Recognition algorithm" patented technology

This invention relates to the field of fault location technology and discloses a fault location method for a long-distance conveying inspection robot. The method includes: when the inspection robot performing long-distance conveying is detected to have reached an entrance equipped with an RFID tag, reading the RFID tag to determine the section of the long-distance conveying route to which the inspection robot belongs, and calculating the distance between the inspection robot and the entrance; acquiring an image of the idler roller taken by a camera mounted on the inspection robot, and using an image recognition algorithm to identify the idler roller code; when a fault signal is received, concatenating the section of the long-distance conveying route to which the inspection robot belongs, the distance between the inspection robot and the entrance, and the idler roller code identification result to obtain the position of the inspection robot in the long-distance conveying, and generating an alarm message. This invention combines RFID and visual recognition to accurately locate the inspection robot's position, improving positioning accuracy and thus improving fault handling efficiency.

The application belongs to the technical field of intelligent security and protection home, and relates to a smart floor system capable of identifying falling behaviors, which comprises a smart pressure sensing floor, a data acquisition system and an upper computer, the smart pressure sensing floor is a composite structure of a solid floor and a flexible sensing fabric, the flexible sensing fabric contains 40 electrodes and 40*40 array distributed sensors, and a piezoresistive sensing layer is constructed by a graphene-carbon black mixed filler to form a conductive network; the data acquisition system collects signals through FPC flat cable, transmits data to the upper computer through a sensor router and a public network router; the upper computer adopts a "cloud-edge-end" architecture, carries a machine learning identification algorithm, can accurately identify falling and intrusion behaviors, the falling identification accuracy is more than 98%, and the response time is less than or equal to 1 second; the terminal side is associated with a cloud program to realize alarm pushing and data viewing functions. The disclosed system simultaneously integrates two core functions of old people falling monitoring and home safety protection, and meets the diversified needs of users living alone.

The application discloses a Helicobacter pylori breath test signal denoising fitting method and system, and belongs to the technical field of medical detection and signal processing. The method processes the continuous monitoring signal of the 13CO2 / 12CO2 isotope ratio in exhalation: firstly, the original signal is subjected to band-pass digital filtering to suppress high-frequency noise; then, low-frequency drift is eliminated through adaptive baseline correction to generate a drift-removed signal; subsequently, a robust peak value recognition algorithm is used to extract physiological response characteristics; finally, the identified peak value is taken as the center, a nonlinear function with physiological constraints is used for local fitting, and the Delta Over Baseline value is calculated based on the generated fitting curve with high precision. The application effectively solves the signal distortion problem caused by breathing fluctuations, instrument noise and the like in in vivo detection, realizes high-fidelity restoration of the Helicobacter pylori metabolic response, and significantly improves the detection accuracy and reliability of the breath test.

Disclosed is a safety shared control system based on performance of teleoperator, including a master teleoperation system, a slave robotic manipulator system and a communication module; where the master teleoperation system includes EEG signal measurement of a teleoperator, hand controller operation input, and upper computer software, and the upper computer software includes a graphical user interface (GUI), safety simulation for protecting the safety of a robot, and a PoT model; and the slave robotic manipulator system includes a robotic manipulator, a vision camera and lower computer software, and the lower computer software includes a target recognition algorithm, an autonomous controller, and a shared controller for dynamically allocating human-robot control weights. A safety control system coefficient of the teleoperator is identified through safety simulation prior to actual operation, such that a safety control method for different operating experience is implemented, and safety operation of a teleoperation robot is realized.

The application relates to an AR glasses binocular calculation algorithm, which is applied to AR glasses provided with left and right light machines, an eye camera and a master control module, and comprises the following steps: S1, initialization and test image projection, after system initialization and ghost image displacement correction are completed, the left and right light machines respectively project binocular calibration images to the eye camera; S2, center calibration positioning, left and right eye field images are acquired through the eye camera, and left eye center coordinates and right eye center coordinates where the binocular calibration images in the left and right eye field images are respectively extracted through an image recognition algorithm; S3, binocular alignment deviation calculation, the position alignment deviation of the left and right eyes is calculated according to the left eye center coordinates and the right eye center coordinates acquired in step S2; S4, dynamic light machine position adjustment, taking the center coordinates of the eye camera as the reference, the left and right light machines are adjusted in the horizontal direction and the vertical direction; and S5, iterative optimization, steps S2-S4 are repeated until the binocular alignment or the maximum iteration number is reached.

A cupholder temperature control system including a pressure sensor for detecting an object within a cupholder, a temperature sensor for detecting a current object temperature of the object, an optical sensor for capturing an image of the object, a processor to perform an image recognition algorithm to detect at least one of a logo, a bar code and a quick response code on the object, to determine the object is a beverage container in response to a detection of the at least one of the logo, the bar code and the quick response code, to generate a control signal in response to the current object temperature differing from an ambient temperature by a threshold amount and the object being a beverage container, and a heat transfer system configured to maintain the current object temperature of the object in response to the control signal.

The application discloses a boride identification method, system and related device, and relates to the field of alloys.The method comprises the following steps: obtaining a metallographic image of a titanium-aluminum alloy; converting the metallographic image into a first image represented by color in terms of hue, saturation and brightness; determining a saturation threshold and a brightness threshold; dividing each pixel point in the first image into a boride category and a background category according to the saturation threshold and the brightness threshold; obtaining a binary image of the first image according to the classification results of the pixel points; identifying at least one target region composed of multiple target pixel points in the binary image by using a continuous region identification algorithm; and determining the target region as a boride region.The application screens the pixel points by using saturation and brightness, so that the possibility of selecting a shallow corrosion structure is reduced; after a suspected boride region is selected, the continuous region identification algorithm is used to identify and screen according to the shape characteristics of the boride. Therefore, the application can effectively improve the identification accuracy of the boride.

We provide the system. [Solution] A processing device that uses a facial recognition algorithm to detect human features from video data and compares them with a subject list, A device that monitors abnormalities in entrances and windows using a vibration detection unit and an opening / closing detection unit, A system that analyzes a person's behavior and facial expressions to quantify the degree of danger and issues a warning based on that evaluation, A means by which the information processing unit shares information about abnormal individuals with multiple external organizations, A means of transmitting a warning to a mobile communication device when an anomaly is detected, A system that includes this.

The application discloses a multi-peak decomposition time-resolved Raman spectrum thermal map interactive analysis method and system, and belongs to the field of spectrum analysis. The method comprises the following steps: continuously or time-divisionally collecting a to-be-measured sample to obtain Raman spectrum data corresponding to different time points; preprocessing the original Raman spectrum data; extracting spectral characteristic peaks by using an adaptive peak value recognition algorithm, and performing characteristic peak integral calculation; obtaining time contribution coefficients of different chemical components or characteristic signals in the time dimension by using a multi-peak decomposition algorithm; generating a Raman thermal map in the time dimension according to the characteristic peak intensity and / or the time contribution coefficients; establishing an index mapping relationship between the position of the Raman thermal map and the corresponding time Raman spectrum data; and responding to the interactive operation of the user on the thermal map and displaying the Raman spectrum curve at the corresponding time point in real time. The application can realize visual display, dynamic monitoring and interactive analysis of time sequence data of Raman spectrum, and improves the efficiency and accuracy of Raman signal analysis of dynamic samples.

The application relates to an explainable language model training method and system for a knowledge-intensive task, and the method comprises the following steps: S1, acquiring original text data; the original text data is a question proposed according to a knowledge-intensive task; S2, inputting the original text data into a preset large language model for training; S3, adopting a hierarchical functional circuit recognition algorithm to recognize a functional circuit in the trained preset large language model; S4, calculating the correlation degree of the functional circuit and the knowledge-intensive task, wherein the correlation degree is a weighted combination based on a selectivity index and an influence index; and S5, outputting an explainable language module, wherein the explainable language module comprises the functional circuit and the corresponding correlation degree. The functional circuit inside the large language model is induced and recognized, mechanism-level transparency is provided, and the method is far superior to existing surface explanation methods.

The application discloses a kind of based on behavior tree and VLA's dual-arm robot task assembly method, specifically includes: when task starts, behavior tree is executed from root node, signal is sent to child node by fixed frequency;Proposition received by parsing perception module is extracted target object and its spatial relationship by conditional node, and three-dimensional information and attribute are obtained using visual recognition algorithm, to judge whether meet the requirement;Action node generates initial path according to motion planning module, trajectory optimization is carried out according to the dynamics constraint of manipulator, and deviation and disturbance in motion process are monitored in real time, timely adjustment is ensured to complete task, wherein user instruction and scene image information are integrated, joint angle and end effector instruction are generated by means of pre-trained multi-modal VLA model, guide dual-arm efficient cooperation, until successfully complete assembly operation.

The utility model discloses a lawn cleaning robot, include: robot main body, mechanical arm structure, lower layer clamp jaw, camera, dustbin, radar, omni -wheel, control unit, GNSS positioning module and IMU, the mechanical arm structure is controlled to a plurality of rudders, makes the mechanical arm structure realize to lift or put down action on the robot main body upper end, and realizes rotating action, is equipped with upper layer clamp jaw in the foremost end of this mechanical arm structure, and this upper layer clamp jaw is controlled to the first rudder, realizes clamping or release, the lower layer clamp jaw is controlled to the sixth rudder, realizes clamping or release, the camera is controlled to the fifth rudder, realizes rotation, has ROS robot operating system and existing identification algorithm YOLOV11 on the control unit deployment.

The application claims a service-aware resource collaborative allocation method, belonging to the field of mobile edge computing resource allocation, which comprises: analyzing task demand characteristics, obtaining task abstract categories combined with task demand characteristics, and mapping them into corresponding demand types. Based on the task demand type, candidate collaborative node clusters are selected through destruction operators and repair operators, then the task and resource matching problem is modeled as a combinatorial optimization problem, the collaborative benefits of candidate node combinations are calculated, and the edge node combination with the highest collaborative benefit is selected. The application mainly considers the problems of task demand fuzziness and task and resource matching imbalance in the computing power network, designs a task demand type identification algorithm and a resource collaborative allocation strategy, selects a reasonable collaborative node cluster, and reduces the time delay while improving the resource utilization.

The application relates to the field of robot simulation and intelligent control technology, and discloses a mobile robot simulation training platform and equipment, which comprises a hardware layer, an algorithm layer and an interaction layer; the hardware layer is used for constructing a three-dimensional simulation environment comprising a mobile robot model capable of carrying a task load, a task scene model and environmental variables; the algorithm layer is integrated with a path planning algorithm, a target recognition algorithm and a motion control algorithm; and the interaction layer provides a parameter adjustment interface. The platform working process is as follows: firstly, the robot and the environment are initialized; then, a target is identified and an optimal path is planned through target detection and path planning; then, the robot is driven to move and real-time obstacle avoidance is performed through motion control and obstacle avoidance; finally, performance evaluation and parameter optimization are performed. Through real parameter modeling and closed-loop simulation training, the application solves the problems of high traditional development cost, long cycle and distorted simulation results, and provides reliable support for the research and development and algorithm verification of mobile robots.

This application discloses a method, apparatus, storage medium, device, and vehicle for recognizing traffic light colors, relating to the field of traffic technology. It addresses the problem of traffic safety issues arising from the inability to identify false detections when using machine learning-based traffic light recognition algorithms. The method includes: acquiring color recognition result pairs of target traffic lights based on traffic light images, where the target traffic lights include turn signals and straight-ahead lights; determining the intersection type of a target intersection by judging whether each color recognition result pair in a first time period is the same color; and, if a preset color recognition result pair exists among the color recognition result pairs acquired after determining the intersection type, identifying the preset color recognition result pair as a false detection result, where the preset color recognition result pair includes color pairs other than those corresponding to the intersection type.

The application discloses a method for visual identification and grabbing control of embodied intelligent robots in a logistics sorting scene, and relates to the technical field of logistics. The application can identify packages that cannot be identified by existing algorithms or have low identification confidence, perform grabbing tests, and analyze the data obtained during the grabbing tests, thereby achieving the effect of analyzing the rigidity, viscoelasticity and friction safety factor of unknown packages. Through the control commands of the experimental robot gripper and the success of grabbing unknown packages with different rigidity, viscoelasticity and friction safety factors, as well as whether the unknown packages are damaged due to grabbing, a second package grabbing decision model is constructed, the success rate of grabbing unknown and special-shaped packages is improved, and the package damage rate is reduced.

This invention proposes a method for multi-target recognition, tracking, and vibration extraction of bridge cable-stayed bridge videos based on unmanned aerial vehicles (UAVs). The method includes: Step 1: Constructing a refined tilted and slender target detection model for bridge cable-stayed bridges based on the YOLOv11 model; Step 2: Proposing a multi-target tracking algorithm that integrates the tilted and slender target detection model and the StrongSORT algorithm; Step 3: Improving the displacement extraction method by combining SIFT / ORB feature point matching and sub-pixel refinement techniques; Step 4: Designing a UAV motion correction algorithm based on variational mode decomposition and time-frequency domain joint screening; Step 5: Constructing a joint working mode analysis algorithm that combines natural excitation technology and random subspace recognition algorithm. This method achieves high-precision extraction of cable-stayed bridge vibration signals and identification of cable-stayed bridge modal parameters, providing technical support for health monitoring of long-span cable-stayed bridges.

The application discloses a kind of stench detection automatic exhaust device and multi-modal intelligent ventilation control method, including the sensing module and ventilation execution host of wireless connection;There is no electrical connection between sensing module and ventilation execution host;Sensing module includes built-in gas sensor, sensing end microprocessor, sensing end wireless communication module and power supply, for collecting environmental gas concentration data and sending;Ventilation execution host includes fan module, main control unit and mains input interface;Main control unit includes main controller, communication module and motor drive circuit;Communication module is used to receive the data sent by sensing module;Main controller is used to run artificial intelligence recognition algorithm, assess odor type and intensity, and generate control instructions according to the evaluation results;Motor drive circuit drives fan module to run according to control instructions.The present application solves the problem of low response lag and energy efficiency in odor control through wireless gas sensing and multi-mode switchable ventilation control.

The utility model discloses a desktop formula wire joint piece scattered silk defect detection device, including the shell body that is located on the desktop, the bottom inner wall of shell body is installed with high definition imaging mechanism, and the front panel on shell body is installed with detection mechanism, the high definition imaging mechanism includes the inclined base fixed mounting on the bottom inner wall of shell body, and the top inclined surface of inclined base is provided with shooting assembly and adjusting assembly, the detection mechanism includes the plug board of inlay formula fixed mounting on the front panel of shell body, and the inboard of plug board is fixedly connected with glass plate through connecting column, and the side fixed mounting of glass plate near plug board has multiple prisms that are circularly distributed, the utility model discloses through high resolution's industrial camera and advanced image recognition algorithm, can accurately identify the tiny defect of wire joint piece, effectively reduces the misjudgment rate, improves product quality, and simple operation simultaneously, and the design structure of desktop formula is compact, and the floor space is small, and is convenient for installation and operation.

The application discloses a kind of based on image data's track area abnormal detection method and device of machine tool, it is related to image recognition technical field.The method includes: using image recognition algorithm, introduce multi-objective optimization algorithm, build machine tool target detection model, and connect to candidate region segmentation engine and space-time correlation filter engine;Video stream of track area is collected, and pretreatment is carried out, and the image data of continuous frame is obtained;The image data of continuous frame is input to candidate region segmentation engine, and the ROI region of each frame of image data is obtained;ROI region is input to machine tool target detection model, and the detection frame of machine tool target is obtained;The detection frame of machine tool target is input to space-time correlation filter engine, identifies stationary effective target, and carries out abnormal detection to stationary effective target.The application solves the problems of low detection accuracy, difficult model deployment and poor anti-interference ability in the prior art.

This invention provides a multimodal fusion-based intelligent assist control method and system for charging gun cables, comprising: acquiring multimodal sensor data; performing attitude calculation on the charging gun based on the multimodal sensor data to obtain the attitude data of the charging gun, and calculating the effective operating force of the user on the charging gun to obtain a comprehensive perception state vector of the charging gun; constructing a user intention feature vector based on the comprehensive perception state vector, and using an intention recognition algorithm to calculate the probability of the user performing a corresponding operation state intention, and using a state machine to calculate the state transition probability of each detected user intention; predicting the user's hand position parameters based on the user intention, and actually acquiring the charging gun's hanging ring position parameters, and constructing a tracking error function based on the predicted user hand position parameters and the charging gun's hanging ring position parameters; generating a control force for a corresponding following mechanism based on the tracking error function, and converting the control force into a driving command for the following mechanism.

The application discloses a motor control double-arm Rubik's cube robot method based on a robot vision recognition algorithm and belongs to the technical field of motor control double-arm Rubik's cube robot methods. In matlab, the starting and ending speed, the maximum speed, the total time, the uniform speed time, the subdivision number of a stepping motor, the subdivision interval, the rotation angle parameter, the specified control curve T type and the S type are defined according to the subdivision number and the rotation angle. The total pulse number is calculated, the total pulse is distributed to the acceleration, uniform speed and deceleration stages, the specified control curve T type and the S type are defined according to the subdivision number and the rotation angle, the total pulse number is calculated, the total pulse is distributed to the uniform speed, acceleration and deceleration stages, the acceleration stage is divided into two sections, namely, the S type, the jerk acceleration is calculated, the speed change is calculated through integration, and the acceleration stage pulse cycle is generated. When the acceleration stage is the uniform acceleration stage, namely, the T type, the acceleration is calculated, the speed linearly grows in each pulse cycle, and the acceleration stage pulse cycle is generated.

The application relates to a bearing fault recognition algorithm suitable for small sample conditions in the field of intelligent manufacturing and industrial fault diagnosis. The whole method process involves vibration signal data acquisition, data cropping, conversion into a frequency domain graph, large-core multi-scale feature extraction, local similarity comparison, global similarity comparison, and fusion of local and global classification stages, and can cope with the real-time performance, high efficiency and deployability of fault recognition under complex working conditions. Among them, the large-core multi-scale feature extraction realizes unified feature extraction modeling; the local module realizes fine-grained feature matching by combining cosine similarity and a multilayer perceptron; the global module models the semantic association between the support set and the query set by using two-dimensional position coding and cross-attention mechanism. Finally, accurate fault classification is realized by fusing the similarity scores of the local and global modules. The method has high accuracy, strong generalization ability and low computational complexity, and is particularly suitable for online diagnosis and edge deployment in industrial scenes.

The purpose of this application is to provide a method and apparatus for acquiring camera parameter information of a PTZ camera device, comprising: acquiring a video stream captured by a corresponding PTZ camera device, wherein the video stream contains one or more target video frames, each target video frame including parameter indication information corresponding to the camera parameter information of the PTZ camera device; determining recognition result information of one or more target video frames in the video stream based on the parameter indication information using an OCR recognition algorithm; and determining the camera parameter information of the PTZ camera device during the video stream acquisition process based on the recognition result information of the one or more target video frames. This application achieves compatibility with all mainstream manufacturers currently on the market, and at the same time, breaks through the direct network limitation between the PTZ camera device and the caller. As long as the video stream can be acquired, camera parameter information can be recognized by OCR, expanding the video application scenarios of PTZ camera devices and improving the user experience.

This invention relates to the field of intelligent security technology, specifically to a multi-scenario plug-and-play intelligent security identification algorithm deployment method, comprising the following steps: constructing a pseudo-trigger constraint set corresponding to the target scenario; generating a scenario-immune deployment body corresponding to the security identification algorithm to be deployed; and implementing pseudo-trigger filtering on the running output to obtain effective identification results; establishing alarm output rules under the target scenario based on the effective identification results, and connecting the effective identification results that conform to the alarm output rules to the formal alarm link; when repetitive abnormal alarms occur in the formal alarm link, locally correcting the corresponding pseudo-trigger constraint set, and synchronously updating the scenario-immune deployment body. Compared with traditional methods that rely on model training or simple threshold filtering, this invention can accurately identify and target structural interferences with stable reproducibility characteristics, such as tree shadow swaying, light reflection, and periodic flickering.

This invention discloses a method and apparatus for real-time video stream image stabilization and multi-magnification clear imaging. It utilizes an intelligent video stream stabilization linkage module to perform real-time stabilization processing on the video stream, dynamically controlling the triggering of the recognition algorithm and achieving smooth front-end display. A video frame imaging clarity processing middleware selects the highest-resolution frame within the same field of view and outputs a clear image in response to shooting requests. A multi-magnification field-of-view switching module enables real-time seamless switching between different magnification fields of view. The real-time video stream image stabilization and multi-magnification clear imaging method and apparatus provided by this invention can automatically and intelligently overcome practical problems such as video jitter and image blur, consistently providing users with clear and accurate recognition results, thereby greatly improving the system's practicality and user experience satisfaction.

This invention discloses an SSVEP brain-computer interface signal processing and CCA classification method, belonging to the field of brain-computer interface technology. Specifically, it includes the following steps: system hardware configuration and data acquisition, data import and channel configuration, bandpass filtering, data segmentation, baseline correction and rereference, downsampling, bad conduction detection and spherical spline interpolation, independent component analysis (ICA) artifact removal, SSVEP dataset construction, and SSVEP power spectral density analysis. This SSVEP brain-computer interface signal processing and CCA classification method constructs a complete end-to-end processing flow from raw EEG acquisition to final classification output. The interfaces of each module are standardized, facilitating engineering integration and secondary development. It employs ICA combined with an automatic artifact recognition algorithm to reduce manual intervention and improve the efficiency and consistency of batch data processing. The CCA algorithm fully utilizes multi-channel spatial information and the harmonic structure of the reference frequency set to achieve high-accuracy SSVEP frequency recognition under unsupervised conditions, adapting to scenarios with densely arranged multi-stimulus frequencies.