184 results about "Identification error" patented technology

The invention relates to the field of computer vision, and particularly provides a table recognition reconstruction method and system, a terminal and a medium, and the method comprises the steps: firstly decomposing a large-size table image into a plurality of overlapped sub-images, and carrying out the table structure detection and OCR character recognition of each sub-image through parallel recognition; then, sub-graph recognition results are integrated through a coordinate mapping and confidence coefficient weighted fusion algorithm, and boundary errors are eliminated; then, automatically distinguishing common cells based on an area clustering algorithm, merging the cells and a header region, and reconstructing a complete table logic structure; further understanding header semantics through a natural language model and repairing identification errors; and finally, realizing intelligent splicing and standardized output of the cross-page table. According to the method, the memory limitation of the traditional OCR technology is broken through, an oversized table can be processed, the recognition accuracy of a complex structure is improved, and the digitization efficiency of professional documents such as financial statements and engineering drawings is improved.

The invention discloses a self-adaptive text extraction method and system based on artificial intelligence, and the method comprises the steps: carrying out the analysis of the document structure entropy of an example document set, quantifying the noise density, geometric distortion degree and background complexity of the example document set, and carrying out the self-adaptive selection of a preprocessing assembly line intensity grade according to the above; dynamically configuring image preprocessing parameters and AI recognition model parameters, and generating a recognition engine instance to output a preliminary recognition text; after regularized coarse screening extraction is carried out based on key field description, a multi-candidate generation strategy is started for low-confidence-coefficient candidate text fragments, a multi-person cooperative verification process is triggered for lower-confidence-coefficient fragments, finally all the fragments are processed through a text standardization module, and structured text extraction information is output. According to the method, accurate adaptation of processing intensity is achieved through document quality quantitative evaluation, the extraction accuracy and system robustness of complex heterogeneous documents are effectively improved through a multi-level confidence coefficient verification mechanism, and the identification error risk caused by image quality fluctuation or rule solidification is reduced.

The invention relates to the technical field of smart power grids, in particular to a detection signal design method and device for multi-injection-point and multi-load-node identification, and the method comprises the steps: calculating the static sensitivity of a plurality of to-be-identified load nodes corresponding to each target equipment injection node in a plurality of target equipment injection nodes based on steady-state operation data, establishing a detection signal optimization model taking a reactive detection signal as a decision variable by taking the total minimum load model parameter identification error of a plurality of to-be-identified load nodes as a target and combining static sensitivity, and introducing reactive energy constraints and signal structure constraints of a plurality of target equipment injection nodes and solving; generating an optimal reactive power detection signal waveform of each target equipment injection node, performing evaluation by using a target evaluation index, when an evaluation result meets a preset constraint condition, injecting the optimal reactive power detection signal waveform into the system, and after a response is generated, performing identification on a load model parameter by using the response, and obtaining an identification result of each load node to be identified.

The invention relates to the technical field of power system modeling and simulation, in particular to a load model identification error analysis method, device and equipment and a storage medium, and the method comprises the steps: obtaining the prediction output and the measurement output of a preset load model, and building a target function based on the error between the prediction output and the measurement output; decomposing the measurement output to obtain an actual output and a noise component of a preset load model; and based on the objective function and the actual output, establishing a linear model between the identification error of the preset load model and the noise component through a first-order approximation method, and performing linear regression according to the linear model to obtain an analysis result of the identification error. Therefore, by constructing the theoretical model of the load model identification error, the problems that the identification error is difficult to predict, the data processing strategy selection lacks theoretical guidance and the like in related technologies are solved, and a theoretical basis is provided for selecting an optimal load modeling data processing strategy.

The invention discloses a line laser anti-welding noise center point extraction and weld joint point identification method, which is applied to the field of welding quality detection and machine vision measurement. The problems of low extraction precision, poor robustness, difficulty in welding seam point positioning and the like are solved for welding spatter, light reflection and background noise influences. According to the technical scheme, the method comprises the following steps: carrying out Otsu adaptive threshold segmentation, connected domain analysis, morphological opening operation and Gaussian filtering denoising; positioning an interference abnormal line and extracting a preliminary central point by a normal centroid method, and further smoothly extracting a detailed central point in an abnormal region by fitting straight lines of 10 front and back central points to assist judgment; and by calculating a second-order difference value, fitting front and back 50 points at the maximum difference value to identify welding seam points, and optimizing a track. The method can be applied to welding seam positioning and contour detection, the accuracy and stability in a complex noise environment are improved, the welding seam point recognition error is smaller than or equal to 0.2 mm, and the actual requirement is met.

The invention discloses a video end-to-end delay accurate measurement method and a system thereof, and belongs to the technical field of video processing, a multicolor frame sequence is generated through a delay box and is output to a matched display, after being acquired by a camera of a tested device, the multicolor frame sequence is displayed at a display end, and a high-speed camera acquires a tested video and transmits the tested video to a signal processing unit. The signal processing unit achieves the technical purpose of achieving full-link high-precision delay measurement from video signal generation to final display on the premise that the structure of detected equipment is not changed through RGB average value comparison, transition frame detection, unique frame sequence identification ID matching and end-to-end delay accurate calculation, frame-level time synchronization is achieved, and the measurement accuracy is improved. Identification errors caused by camera exposure change, ambient light interference and display color deviation can be effectively suppressed, so that the measurement stability and accuracy are remarkably improved.

The invention discloses a solid tire quality detection method based on infrared and deep learning, and relates to the technical field of tire quality detection, and the method comprises the following steps: in the high-speed operation process of a solid tire, collecting a current infrared image, and extracting a high-gradient thermal texture reference point based on a phase correlation operator; and based on the extracted hot texture reference point, mapping a reference point track in a previous frame of image to the current image through adaptive optical flow tracking, and generating a hot texture original displacement vector field. According to the method, real inter-frame displacement identification is realized through hot texture reference point extraction and optical flow tracking, high-frequency disturbance is suppressed in combination with frequency domain filtering, image position reference is unified through weighted fusion and standardized offset mapping, and meanwhile, low-quality frame input is effectively judged and controlled by introducing continuous verification and dynamic weight regulation. The consistency and robustness of image splicing are integrally improved, the defect recognition error is remarkably reduced, and the detection stability and accuracy of the deep learning model in the industrial environment are guaranteed.

The invention provides a voice robot voice receiving, recognition and error correction method, which comprises the following steps: constructing an intention stability sensing unit, and realizing dynamic judgment on the state of a deliberate map in combination with multi-granularity semantic coding and sliding window variance analysis; when it is detected that the intention is unstable or multiple intentions compete, a parallel intention matching mechanism is adopted, and the accuracy of intention optimization is improved through confidence, semantic continuity and activation frequency fusion; through context significance weight and pruning strategies, historical semantic information is effectively managed, low-contribution content is compressed, and resource overhead is reduced; gating network parameters are continuously optimized on line, the intention recognition effect is improved in real time, the continuity and stability of dialogue intention modeling are enhanced, and the recognition accuracy and the system resource efficiency are both achieved.

The utility model relates to the technical field of wafer thinning machines, in particular to a wafer thinning machine front face and back face recognition mistake proofing mechanism which comprises a base, four limiting holes are symmetrically formed in the top end of the base, electric telescopic rods are fixedly connected into the limiting holes, the top ends of the electric telescopic rods are fixedly connected with a fixing base, and the fixing base is fixedly connected with a wafer thinning machine front face and a wafer thinning machine back face recognition mistake proofing mechanism. A limiting groove is formed in the fixing base, a spring is fixedly connected to the interior of the limiting groove, and a feeler lever located in the limiting groove is fixedly connected to the bottom end of the spring. Then the electric telescopic rod drives the fixing seat and the wafer to descend, the wafer is ground and thinned through the rotating grinding disc along with contact of the wafer and the grinding disc, and under the action of limiting of the fixing ring and the size of the grinding disc, a frame is left at the edge of the ground wafer; the front and back surfaces of the wafer can be easily judged through the frame left at the edge of the wafer.

The invention mainly discloses a quality detection cooperative interaction suite group, which is used for performing defect detection on a continuous material conveyed by a roll-to-roll mechanism. Particularly, the quality detection cooperative interaction suite set further provides the following functions: (1) intuitively displaying the position of the detected flaw on the continuous material; (2) when the flaw detection result is misjudged, a user (a quality inspector) operates a tablet computer provided by the quality detection collaborative interaction suite group to delete the misjudged flaw detection result online; (3) when a flaw category identification error occurs in a flaw detection result, the user operates the tablet computer to correct the flaw category identified by the error online; and (4) when missing detection occurs, the user operates the tablet computer to add one or more images containing flaws into the database.

This application discloses a speech recognition method, apparatus, electronic device, medium, and program product, belonging to the field of audio technology. The method includes: acquiring an identification error corresponding to a secondary path of an audio device, the identification error being used to characterize whether the signals collected by at least two microphones in the audio device are human voice signals; if the identification error is greater than or equal to a first threshold, acquiring a first spectral ratio, the first spectral ratio being the ratio of the spectra of the signals collected by at least two microphones; if the first spectral ratio is greater than or equal to a second threshold, determining that the signals collected by at least two microphones were emitted by a user wearing the audio device.

This invention relates to the field of wind tunnel aerodynamic measurement technology for hypersonic vehicles, and discloses a method for evaluating and optimizing the dynamic load testing accuracy of a wind tunnel force measurement system. The evaluation method involves: equating the force measurement system to a multi-degree-of-freedom spring and mass model; constructing a dynamic differential equation and performing a Laplace transform; then, combining the dynamic stiffness matrix to construct a frequency response function matrix; defining the output of the force measurement channel of a rod-type strain balance and solving for the transfer function from the excitation point to the force measurement channel; combining the reference transfer function and the actual load, calculating the identified load and performing a Fourier transform to obtain the relative identification error of the force measurement system at the excitation point; finally, calculating the power spectrum weighted average identification error at the excitation point and comparing it with the engineering allowable error to determine whether the system has undergone stiffness optimization. This invention not only accurately identifies the dynamic load of the force measurement system but also achieves stiffness optimization, improving the accuracy of unsteady aerodynamic force measurement.

This invention provides a truck model and axle count information comparison and export transaction system, relating to the field of intelligent transportation technology. It includes a multi-source data acquisition unit; an information processing and verification unit, communicatively connected to the multi-source data acquisition unit, for processing the multi-source data based on an artificial intelligence recognition model and constructing a dynamic digital twin model of the vehicle; a multi-dimensional cross-verification mechanism; and a hierarchical decision execution unit, communicatively connected to the information processing and verification unit, for executing hierarchical response operations based on the risk score. By introducing multi-source data fusion and artificial intelligence analysis, this invention significantly improves the accuracy of vehicle model and axle count verification, effectively reducing tolling errors caused by identification mistakes. Through the multi-dimensional cross-verification mechanism, the system achieves real-time comparison and risk warning of vehicle information at entrances and exits, enhancing the real-time identification and prevention capabilities for cheating behaviors such as modified vehicles and hidden axles.

The application discloses a kind of neurogenetic disease identification model construction method and system, it is related to neurogenetic disease identification technical field.The method includes: obtaining multiple identification results and multiple user feature sets of neurogenetic disease identification, and multiple inter-class difference degrees are obtained by analysis;Based on multiple identification results and multiple user feature sets, train first identification model;According to multiple inter-class difference degrees, configure user feature synthesis strategy, generate synthetic user feature set and synthetic identification result, carry out the reinforcement learning of first identification model, obtain second identification model;Test the combined identification error rate of second identification model, combined with multiple inter-class difference degrees again configure synthetic strategy, continue to generate synthetic user feature set and synthetic identification result, carry out reinforcement learning, obtain third identification model.The application solves the problem that neurogenetic disease identification model is poor in generalization performance, easily confused class and low in recognition accuracy under small sample condition.

The invention relates to the technical field of intelligent data terminal remote control, in particular to a bar code data transmission method allowing missing codes, which comprises a code generation step of data file conversion and a decoding step of scanning equipment. When data transmission is carried out, a to-be-transmitted data file is coded to generate a series of bar codes, continuous scanning is carried out through a code scanning device, the data is completely transmitted into the device, a fault-tolerant mechanism is constructed by using an RS error correction algorithm principle, RS error correction code words are added to realize redundancy, and part of the bar codes cannot be successfully scanned or have recognition errors. And the system can still utilize the redundant information reserved in the error correction group in the subsequent decoding stage to recover missing or wrong code words, so as to reconstruct a complete data file. Compared with the prior art, the code scanning equipment does not need to interact with equipment for displaying bar codes, has higher stability, is particularly suitable for a data transmission scene in a dynamic or interference environment, and has higher feasibility and adaptability.

The application provides a kind of steam turbine governing system electric power step response simulation error analysis method and device, it is related to electric power equipment performance detection field, including: according to the parameter sample of steam turbine governing system electric power step response, preset sample polynomial is fitted, corresponding peak power arrival time is obtained;According to the sample polynomial and historical sample of fitting, the starting step time and steady state entering time of the parameter sample are determined;According to the peak power arrival time, the starting step time and steady state entering time determine the parameter identification error of the steam turbine governing system electric power step response.The application can determine the parameter identification error of steam turbine governing system electric power step response.

This invention provides a diesel engine abrasive particle color recognition method based on adaptive clustering and a dual-mapping standard, comprising the following steps: automatic extraction of diesel engine abrasive particle color through adaptive clustering; and automatic recognition of diesel engine abrasive particle color based on the extracted abrasive particle color using a dual-mapping standard. The beneficial effects of this invention are: automatically extracting abrasive particle color and associating it with abrasive particle material and harsh wear process, satisfying the two major work objectives of diesel engine abrasive particle material identification and harsh wear process identification, and achieving automatic recognition of diesel engine abrasive particle color using a dual-mapping standard; introducing a standard color system, the extracted color features are first compared with the standard color system to determine the color name, and then compared with an abrasive particle color library based on the standard color system for identification, which can effectively reduce the impact of uncertain contaminants such as dust and fibers on the identification process and reduce identification errors.

The application relates to the technical field of multi-agent systems, and provides a multi-agent sensor fault diagnosis method and system based on adaptive learning, which comprises the following steps: acquiring control signals and outputs of each multi-agent subsystem; based on the control signals and the outputs, an adaptive observer is used to obtain a state vector of the adaptive observer; by comparing the state vector of the adaptive observer with a state vector of the multi-agent subsystem, fault information of a sensor corresponding to the multi-agent subsystem is obtained; wherein the adaptive observer ensures that a tracking error and a weight vector error are bounded, the tracking error is a difference value between an output estimation of the multi-agent subsystem and an output of the multi-agent subsystem, and the weight vector error is a difference value between an estimated weight of a neural network in the adaptive observer and an ideal value; and the final consistent boundedness of identification errors is ensured.

The invention discloses a material standardization method and device, and relates to the technical field of computers. A specific embodiment of the method comprises the steps of obtaining material description content, and determining at least one input type corresponding to the material description content; aiming at each input type, determining type content of the material description content aiming at the input type; extracting a type text corresponding to the type content; combining the type texts corresponding to the input types to generate a combined description text corresponding to the material description content; extracting at least one attribute key value pair from the combined description text; wherein the attribute key value pair comprises an attribute name and an attribute value; and according to the at least one attribute key value pair, matching a standard material corresponding to the material description content from a standard material library. According to the embodiment, the material description content can be subjected to standardization processing, the problem of material identification errors is reduced, and the product quality and equipment safety of enterprises are guaranteed.

A control unit for improving an annotation system in an intelligent model is provided. The control unit 10 receives data from a plurality of data sources and fine-tunes the smart model 14 using the received data. The control unit 10 then determines a prediction / tag for each data point of the received data and identifies, by an error analysis module 16, at least one error in the determined prediction / tag for each data point of the received data. The control unit 10 transmits samples with identified errors back to the data set via a sampling policy module 18 for fine tuning the smart model 14, such that the annotation system 12 in the smart model 14 is improved.

The application provides a tunnel blasting charge structure identification recording method, device, equipment and medium, comprising: acquiring a blasting signal set generated by a wireless vibration measuring unit during tunnel peripheral charge structure blasting; processing the blasting signal set to generate multiple sets of dimension reduction data; calling a trained KNN model to process the dimension reduction data to generate a charge structure type corresponding to the blasting signal set, wherein the KNN model automatically records vibration data corresponding to the charge structure type generated during tunnel blasting. In addition, most of the methods on the market for identifying and recording charge structure types and vibration data according to vibration characteristics are manually recorded and supervised in real time, which is labor-intensive and prone to omissions, errors, type identification errors and other problems, which may affect the subsequent tunnel blasting excavation vibration reduction and loss reduction effect, thereby reducing the self-stability of the surrounding rock.

A kind of operating condition under rotating electric machine inherent frequency calculation method discloses a kind of operating condition under rotating electric machine inherent frequency acquisition method, belong to motor structure modal parameter online identification technical field.Method includes: in stator tooth portion uniform distribution N≥4 one-way vibration sensor;Motor is under load and vibration signal is synchronously collected;Signal zero-phase band-pass filtering;The correlation function of reference point and response point is calculated, and complex modal modeling is complex exponential series;Through single-sided mutual power spectrum transformation and multiple starting time sampling matrix equation is constructed, coefficient vector is obtained by SVD solution;Root is sought by coefficient and constructs characteristic polynomial, mapping obtains system continuous pole, and then simultaneously output inherent frequency and damping ratio.The whole process does not need to stop, does not need to add excitation, noise resistance is strong, frequency identification error is less than 1%, provides modal data under real boundary condition for motor NVH optimization and structural modification.

The application provides a multi-layer assembly type building prefabricated component hoisting construction method, the method is deeply fused with a special hoisting coordinate system and a BIM model, three-axis high-precision positioning of a lifting hook and accurate mapping of a component coordinate are realized, combined with posture stability, dynamic obstacle avoidance and millimeter-level calibration technology, the problem of cross-floor space interference is solved, hoisting precision and overall efficiency are greatly improved, a three-dimensional multi-dimensional identification system is adopted, identification errors are completely eliminated, automatic identification and time consumption statistics of all process nodes are realized, hoisting sequence early warning and three-dimensional visual progress display are matched, the whole construction process is controllable, and the problems of low coordinate positioning precision, disconnection of multi-platform data and unquantifiable work efficiency in the prior art are solved.

The invention provides a vehicle welcome control method and a vehicle, and relates to the technical field of vehicle control, and the method comprises the steps: obtaining a terminal identity label of a user terminal in response to the detection that the user terminal of a current user is connected to a wireless charging module of the vehicle, and the terminal identity label of the user terminal passes verification; and then, according to the terminal identity label, determining welcome configuration information of the current user, and controlling the vehicle to execute at least one welcome action through the welcome configuration information, so as to trigger vehicle welcome under the condition that the user terminal is connected and the wireless charging module passes verification, thereby realizing welcome triggering based on the wireless charging module. The welcome response efficiency is ensured while the welcome cost is ensured, and recognition errors caused by environmental factors or sensor errors are avoided, so that the welcome reliability is ensured; moreover, the method can also obtain the welcome configuration according to the terminal identity identifier, so that the welcome effect better meets the user demand, and the user experience is further improved.

The present application relates to a kind of methods for coordinating left and right light controller work using CAN message, comprising the following steps: in the first time window after the power-on of left controller and right light controller, local pin level is detected and bus message is listened, and it is determined that itself is left light controller or right light controller by level state;In the first time window, left controller or right light controller listens CAN bus, whether the state message of non-controller side is received is judged, if so, continue normal work;Otherwise, trigger identification error flag.The method for coordinating left and right light controller work using CAN message, device, processor and its computer readable storage medium of the present application are adopted, the pain points of left and right light controller identity misjudgment and playing out of sync are solved, single point failure caused by wire harness open circuit or unreliable connection is avoided, the function safety level of turn signal system is greatly improved, and the defect that traditional time reference synchronization easily causes software competition and then reduces reliability is overcome.

Disclosed is a robot. One or more processors, when instructions are individually or collectively executed, cause the robot to: monitor at least one function of the robot related to a preset task; when an error related to the at least one function is identified, identify the type of the error; identify, on the basis of the type of the error, the degree of feasibility of the preset task after the error occurs; and change the at least one function on the basis of the degree of feasibility of the preset task.

We provide the system. [Solution] A data processing means for analyzing user answer information and identifying errors, A data analysis means using generation techniques to analyze the cause of the identified error, Information gathering means that automatically collects information related to user errors from external information resources, Information display means for providing collected information to the user, A communication means for transmitting answer information to a data server via a communication network, A means for receiving analysis results from a server and presenting them to the user, A system that includes this.

The application discloses a battery swap station operation management method for confirming whether a to-be-battery-swap vehicle entering a battery swap station is an authentication vehicle, and the method comprises the following steps: acquiring frame number information of the to-be-battery-swap vehicle; acquiring license plate number information of the to-be-battery-swap vehicle; determining whether the frame number information and the license plate number information are in an authentication corresponding relationship; if the frame number information and the license plate information have the authentication corresponding relationship, authentication is successful, and it is confirmed that the to-be-battery-swap vehicle is an authentication vehicle; and if the frame number information and the license plate number information are in a non-corresponding relationship, authentication fails, and prompt information is output. Therefore, the application can ensure the reliability of to-be-battery-swap vehicle information, avoid identification errors, especially greatly avoid abnormal battery swap caused by license plate input errors and identification errors, and greatly improve service reliability.

The invention mainly discloses a cloth inspecting machine integrated with a quality detection cooperation and interaction suite group. The cloth inspecting machine is integrated with the quality detection cooperation and interaction suite group. Particularly, the quality detection collaborative interaction suite group further provides the following functions: (1) providing and intuitively displaying the position of the detected defect on the cloth; (2) when the flaw detection result is misjudged, a user (a quality inspector) operates a tablet computer provided by the quality detection collaborative interaction suite group to delete the misjudged flaw detection result online; (3) when a flaw category identification error occurs in a flaw detection result, the user operates the tablet computer to correct the flaw category identified by the error online; and (4) when missing detection occurs, the user operates the tablet computer to add one or more images containing flaws into the database.

Arrangements for quantum enabled error identification are provided. A computing platform may train an artificial intelligence (AI) engine. The computing platform may extract metadata from one or more interconnected systems. The computing platform may use the metadata to map connections within the one or more interconnected systems. The computing platform may generate one or more dynamic markers. The computing platform may convert the metadata into qubits. The computing platform may generate one or more quantum enabled differential privacy components. The computing platform may inject the components into the one or more interconnected systems. A dynamic marker may be highlighted. An error may be identified based on the dynamic marker that was highlighted. An action may be identified based on the error. The action may be automatically executed by the computing platform.