Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

95results about How to "Reduce recognition errors" patented technology

Unmanned aerial vehicle accurate position landing method and system based on image recognition analysis

The invention discloses an unmanned aerial vehicle accurate position landing method and system based on image recognition analysis, particularly relates to the technical field of aircraft landing, and aims to solve the problem of influence of jitter interference caused by micro airflow on landing precision. Identifying the jitter signal intensity and judging whether the jitter signal intensity is abnormal or not by detecting the posture change of the unmanned aerial vehicle in the landing process in real time; when the jitter is abnormal, carrying out multi-frame image superposition processing to generate a stable image sequence; detecting fine texture disturbance between adjacent frames based on multi-resolution fractal dimension analysis, and compensating non-linear influence caused by micro airflow; recording a jitter trajectory and calculating the relative displacement of the landing target through reverse trajectory reconstruction; extracting calibrated landing area features in combination with a convolutional neural network, generating an offset compensation value of a landing point, and dynamically correcting a landing path; according to the invention, the landing precision of the unmanned aerial vehicle in a complex environment is improved, and reliable guarantee is provided for automatic landing of the unmanned aerial vehicle.
Owner:JIANGXI PROVINCIAL MILITARY & CIVILIAN INTEGRATION RES INST

High-flow-resistance multi-point observation type fish school monitoring underwater robot based on multi-source data fusion

The invention discloses a high-flow-resistance multi-point observation type fish school monitoring underwater robot based on multi-source data fusion. The high-flow-resistance multi-point observation type fish school monitoring underwater robot comprises a shell, a self-adaptive ocean current resistance system and an energy system. The self-adaptive ocean current resistance system comprises an ocean current information detection module and a PID control algorithm real-time regulation and control propeller module; the ocean current information detection module is used for collecting ocean current information in real time, including water velocity, direction and turbulence intensity; the control algorithm real-time regulation and control thruster module regulates and controls the power and direction of a thruster in real time by utilizing a PID control algorithm according to ocean current changes, and it is ensured that the underwater robot keeps stable in complex ocean currents; the telescopic multi-sensor fusion system collects multi-dimensional data of a fish school by combining a plurality of sensors, and performs fusion analysis on the data by using a deep learning algorithm, thereby realizing high-precision fish school identification and tracking. And an integrated ocean current-solar energy system enables the underwater robot to have long-time cruising ability.
Owner:SANYA YAZHOU BAY INST OF DEEP SEA SCI & TECH SHANGHAI JIAOTONG UNIV +1

User information input method and system in intelligent television all-in-one machine

The invention relates to the technical field of television image communication, in particular to a user information input method and system in an intelligent television all-in-one machine. The method comprises the following steps: collecting a user image, carrying out white balance correction and focusing processing, generating a moving object focusing image, carrying out identification and attention tendency analysis on a character in the focusing image, obtaining character tendency data, analyzing a historical interaction image, identifying an interaction mode and constructing a user intention prediction model, and obtaining a user intention prediction result. And when the tendency data shows that the user preferences the television content, intention drilling and auxiliary adjustment are performed on the current image based on the prediction model, predicted user input information is generated, logic conflicts in the predicted information are detected, feedback optimization is performed, and final user input information is output. According to the invention, a more accurate and more flexible user information input method is realized.
Owner:SHENZHEN MEIJIA TECH CO LTD

Tooth position and tooth disease automatic detection method based on YOLOv8

The invention provides a tooth position and tooth disease automatic detection method based on YOLOv8. The method specifically comprises the following steps: collecting a panoramic curved surface fault piece and carrying out standardization processing; labeling tooth positions and tooth disease conditions in the image by using professional software to generate a labeling data set; a C2fConvNeXtv2 module is integrated in the YOLOv8 backbone network, so that the feature extraction capability is enhanced; an ASPP module is adopted to replace an SPPF module so as to optimize capture of multi-scale features; the Focal-EIoU Loss is introduced to improve the regression precision of the bounding box; training an improved YOLOv8-Teth model on the tooth disease data set, and selecting an optimal model as a final tooth disease detection model; and inputting a to-be-detected panoramic curved surface fault into the trained model to obtain an automatic identification result of the tooth position and the tooth disease condition. The model solves the problem of low target detection precision caused by high similarity and complex texture of teeth in tooth detection, provides powerful technical support for early diagnosis and treatment of dental diseases, and improves clinical diagnosis efficiency and accuracy.
Owner:SHENYANG INST OF COMPUTING TECH CO LTD THE CHINESE ACAD OF SCI

Loess landslide detection method and system based on grade differentiation learning grey wolf optimization

The invention belongs to the technical field of image processing, and discloses a loess landslide detection method and system optimized by a multi-population grey wolf optimization algorithm based on grade differentiation learning. Performing data enhancement on the landslide data set by using a data enhancement generator; designing a grey wolf optimization algorithm model based on multi-level differential learning through the segmentation network model, and optimizing hyper-parameters in the segmentation network model by using an improved algorithm to realize adaptive adjustment and cooperative calculation of three hyper-parameters, namely a learning rate, an iteration number and a batch processing number, in the segmentation network; and finally carrying out loess landslide detection. According to the method, algorithm improvement is carried out, so that the balance capability of global exploration and local development is enhanced, and later diversity loss and local optimum are avoided. By combining the excellent semantic segmentation capability of VIG and the improved grey wolf optimization algorithm, the method can effectively overcome the feature extraction difficulty caused by factors such as geological structure and vegetation coverage, and significantly improves the accuracy of loess landslide recognition.
Owner:NORTHWEST UNIV

Fresh cut rose flower openness grading method

The invention relates to a fresh rose cut flower openness grading method, which comprises the following steps: acquiring a cut flower image, and generating a data set after preprocessing; labeling a data set and carrying out visual analysis, and carrying out data expansion on a malformation category image of the data set; constructing a neural network model for feature extraction and classification, training and optimizing the neural network model, inputting the data set into an optimization model, obtaining a prediction result output by the optimization model, determining confidence and bounding boxes in the prediction result, sorting the bounding boxes based on the confidence, screening the bounding boxes through a reference box, and outputting a recognition result. The method can better adapt to the inconsistency of rose features in the openness grading of the fresh cut roses, improves the recognition accuracy of the network, achieves the quick and accurate positioning and classification of the openness recognition parts of the fresh cut roses, automatically achieves the openness grading of the fresh cut roses, reduces the recognition errors caused by manual intervention, and improves the recognition efficiency. The openness identification efficiency and accuracy are improved; and the enterprise profit is improved.
Owner:KUNMING UNIV OF SCI & TECH

Hard disk backboard and server

The invention discloses a hard disk backboard and a server, and relates to the technical field of servers. The power supply control interface of the backboard controller is connected with the hard disk connector, the power supply control function of the hard disk is fully utilized to control the work of the hard disk, the situation that the occupied area of the backboard is additionally increased in an electronic fuse mode is avoided, and connection is achieved only in an interface mode; meanwhile, a clock buffer is added and integrated with a power supply control function of the hard disk to realize a power-on time sequence which is strictly required by a high-speed signal of a specific hard disk, so that the hard disk identification error problem caused by a time sequence error risk is reduced, the hard disk identification accuracy is improved, and the adaptation compatibility between the hard disk and a mainboard is correspondingly improved. Besides, the backboard controller and the hard disk connector are connected through the first communication interface, the situation that key information is leaked due to the fact that conventional hard disk information acquisition occupies a polling communication protocol for a long time is avoided, and the safety of hard disk information acquisition is improved.
Owner:INSPUR SUZHOU INTELLIGENT TECH CO LTD

Zone area topology identification method and system based on Kirchhoff's law

The invention discloses a transformer area topology identification method and system based on Kirchhoff's law, and relates to the technical field of low-voltage transformer area topology identification, and the method comprises the steps: obtaining low-voltage rural power grid transformer area LMUs and user meter data, and carrying out the preprocessing; for LMUs data, establishing an LMUs-LMU identification relation model based on the Kirchhoff's law; based on the sequence of the LMUs from high to low, establishing an LMU-LMU relation model based on the Kirchhoff's law for the LMUs to be compared and the LMUs in the existing LMU-LMU topological structure one by one, and obtaining the real positions of the LMUs to be compared in the LMU-LMU structure; establishing an LMU-user meter identification relation model; and based on the identified complete LMU-LMU structure, establishing an LMU-household meter identification model based on Kirchhoff's law for each household meter and all LMUs, and finally determining the LMU to which the current household meter belongs. According to the invention, in a power topology identification scene based on data analysis, the investment of resources such as manpower and material resources is reduced, the topology identification accuracy can be ensured, and a set of transformer area topology identification system with low cost and high efficiency is formed.
Owner:YUNNAN POWER GRID CO LTD

Multi-level low-voltage transformer area topology identification method based on linear regression and related system

The invention belongs to the field of low-voltage power distribution of an electric power system, and discloses a multi-level low-voltage transformer area topology identification method based on linear regression and a related system. The connection relation between each meter box and a branch node, a main line and a transformer in a low-voltage transformer area can be identified more accurately, and identification errors are reduced. According to the method, the complex connection relation between the meter box and the branch nodes can be effectively processed and analyzed by using the multi-scale similarity measurement and hierarchical clustering method, and the problem that the traditional method is difficult to identify when facing a complex topological structure is solved. According to the invention, by automatically acquiring and processing the voltage load data, the current data and the correlation information thereof, manual intervention can be greatly reduced, and the efficiency of topology identification is improved.
Owner:CHINA ELECTRIC POWER RESEARCH INSTITUTE CO LTD +1

Special field text semantic element recognition method based on enhanced CRF layer

The invention relates to the field of text semantic analysis, in particular to a special field text semantic element recognition method based on an enhanced CRF layer, and the method comprises the steps: carrying out the sample screening analysis of a sample library, so as to obtain a plurality of training samples; for a single training sample, performing semantic element extraction on the training sample according to a preset strategy function to obtain a tag sequence; calculating an accumulated discount reward corresponding to the tag sequence; obtaining a probability award expected value corresponding to the tag sequence by adopting a preset objective function, performing gradient calculation by adopting a preset gradient formula, and performing parameter updating for the model parameters; determining whether to reselect the label according to a preset exploration strategy; under a preset stopping condition, stopping model training learning; according to the method, the capability of effectively capturing the complex dependency relationship between the tag sequences by the model is improved, so that the accuracy and robustness of the model on a semantic element recognition task are improved, and the capability of processing a long text is improved.
Owner:NO 63921 UNIT OF PLA

Power grid station wiring fault auxiliary analysis system

The invention relates to a power grid station wiring fault auxiliary analysis system, belongs to the technical field of fault analysis, and solves the problem of low efficiency of existing wiring fault analysis. Comprising a station drawing importing module used for importing an original image of a power grid station drawing, segmenting the original image to obtain each wiring diagram, and obtaining a to-be-detected image and an auxiliary image based on each wiring diagram; the wiring data identification module is used for identifying elements and text contents by utilizing a plurality of neural network models according to each to-be-detected image and an auxiliary image thereof, and associating the elements and the text contents; the wiring topology identification module is used for identifying each line segment in each to-be-detected image and constructing wiring link information according to elements connected with the line segments or the line segments; the auxiliary analysis input module is used for selecting a power grid station drawing, a wiring diagram, an element mark number or a wire number; and the fault auxiliary analysis module is used for analyzing a fault source link and an influence link from the wiring link information according to the selected component mark number or wire number. The fault analysis efficiency is improved.
Owner:STATE GRID SHANGHAI MUNICIPAL ELECTRIC POWER CO

Non-woven fabric surface defect detection device, movable pressing assembly and pressing method thereof

The invention belongs to the technical field of non-woven fabric detection, and discloses a non-woven fabric surface defect detection device, a movable pressing assembly and a pressing method.The non-woven fabric surface defect detection device comprises a bearing assembly, a heat dissipation assembly is connected to the bearing assembly, the bearing assembly comprises a supporting shell, a carrying cavity is formed in the supporting shell, and a movable pressing assembly is arranged in the carrying cavity. A carrying cavity is formed in the supporting shell, a recognition assembly is connected to the interior of the carrying cavity, a moving assembly is arranged on the supporting shell in a penetrating mode, non-woven fabric is connected to the moving assembly, the recognition assembly can recognize the non-woven fabric, and through mutual cooperation of the recognition assembly and the working assembly, the non-woven fabric can be recognized; the non-woven fabric defect detection device has the advantages that the non-woven fabric defect detection accuracy is improved, the non-woven fabric moving path is restrained by the working assembly, the surface stretching performance of the non-woven fabric when the non-woven fabric passes through the identification assembly is guaranteed, the identification error of the identification assembly is reduced, and the detection efficiency of the defect detection device is improved.
Owner:ZHEJIANG FULIBANG NON WOVEN TECH CO LTD

On-orbit assembly robot ground test system and method based on air floating platform

According to the on-orbit assembly robot ground test system and method based on the air floating platform, the air floating platform serves as a support, gravity unloading is provided through two air floating simulators floating on the platform, and a mechanical arm installed on one air floating simulator picks up the other air floating simulator; and enabling the large-tolerance butt joints which are respectively arranged on the flexible simulation piece cross beams of the two air floatation simulators to be close to each other until the butt joint is completed, so as to complete the verification of the related auxiliary butt joint capability of the space manipulator. In the process, the vertical rod of the flexible simulation piece is connected with the air floating simulator at the mounting end of the mechanical arm, and flexible damping simulation in the horizontal direction is provided. Different flexible simulation pieces are matched with the air flotation device, vertical gravity unloading, horizontal controllable force and torque control and flexible base damping disturbance in the mechanical arm actuation process are met at the same time, the in-orbit working condition of the space mechanical arm can be fully simulated, and the auxiliary butt joint capacity of the mechanical arm is verified.
Owner:CHINA ACAD OF AEROSPACE SCI & TECH INNOVATION

Work piece surface character recognition method and device, computer equipment and storage medium

The invention relates to a workpiece surface character recognition method and device, computer equipment, a storage medium and a computer program product. The method comprises the steps of obtaining a to-be-recognized image shot for a placement station where a workpiece is located, performing region division on the to-be-recognized image, determining a character region containing character information from the to-be-recognized image, and screening out a non-character region, thereby improving the accuracy of determining the character information. Affine transformation processing is carried out on the area boundary of the character area to obtain the target image area, the posture of the character information is corrected, and the accuracy of determining the character information is improved. And performing character extraction on the target image area to obtain the character information on the surface of the workpiece, and by adopting the method, performing affine transformation processing on the area boundary of the character area to obtain the target image area, the identification error caused by character deformation is reduced, and the accuracy of determining the character information is improved.
Owner:SPEEDBOT ROBOTICS CO LTD

Gesture recognition method, terminal device, storage medium and computer program product

The invention discloses a gesture recognition method, terminal equipment, a storage medium and a computer program product, and relates to the technical field of signal processing, and the method comprises the steps: obtaining a pre-trained basic gesture recognition model; obtaining user training data of the target user, and adjusting a preset offset calibration module based on the user training data to obtain an individual calibration module; a target gesture recognition model is constructed based on the individual calibration module and the basic gesture recognition model, and when gesture recognition is conducted on the to-be-recognized electromyographic signals of the target user according to the target gesture recognition model, feature vectors corresponding to the to-be-recognized electromyographic signals are adjusted according to the individual calibration module. The individual calibration module is introduced to capture and adapt to the unique features of the target user, so that the target gesture recognition model can better capture and adapt to the unique features of the target user, recognition errors caused by individual differences are reduced, and the accuracy of gesture recognition is improved.
Owner:WEIFANG GOERTEK ELECTRONICS CO LTD

Multi-commodity automatic correction and identification method and intelligent self-service cash register

The invention relates to the technical field of image recognition, and discloses a multi-commodity automatic correction and recognition method and an intelligent self-service cash register, and the method comprises the steps: extracting a commodity front image from a commodity image; identifying label areas, and selecting associated feature points for each label area; segmenting a commodity area from the front image of the commodity; determining an associated commodity area of each label area; extracting a commodity side image, and determining an associated commodity area of each label area in the commodity side image; reconstructing a three-dimensional model of the commodity; scanning a label area to obtain commodity information, and verifying the commodity information based on a gravity sensor and the three-dimensional model of the commodity; according to the invention, the accuracy of commodity identification and the settlement efficiency are improved, the user experience is improved, the security and reliability of the system are enhanced, and a more efficient, convenient and reliable self-service cashier solution is provided.
Owner:WUXI PINGUANG IOT TECH CO LTD

A micro-expression recognition method, system, device and storage medium based on multivariate optical flow feature fusion

The present invention discloses a micro-expression recognition method, system, device and storage medium based on multivariate optical flow feature fusion. The method includes obtaining original video, generating optical flow images respectively through multiple optical flow algorithms, extracting horizontal optical flow vectors and vertical optical flow vectors therein for superposition, obtaining horizontal images, vertical images and overall images, and inputting them into the MRKCBN network model for feature extraction, outputting corresponding image output features to obtain local features; inputting the optical flow image generated by one of the optical flow algorithms into the VIT network model for feature extraction and outputting global features, splicing local features and global features to obtain splicing features; fusing the splicing features through the APSB module to obtain fused features; and classifying the fused features through the fully connected layer to obtain micro-expression recognition results. The present invention uses multiple optical flow algorithms to fuse optical flow information to improve the accuracy of recognition; and takes into account both local features and global features through MRKCBN and APSB.
Owner:NANJING XIAOZHUANG UNIV

Medicine box medicine taking method and medicine box storing and taking device based on image model matching recognition

The invention relates to a medicine box medicine taking method and a medicine box storing and taking device based on image model matching identification, the medicine box medicine taking method and the medicine box storing and taking device based on the image model matching identification are ingenious in design, key feature values on multiple faces of a medicine box are extracted and learned to be bound with the medicine box, and the key feature values are uploaded to a medicine box database; then the medicine boxes with high similarity are placed in different trays according to different specifications, the recognition error can be reduced, the medicine boxes are shot through the first camera, and the medicine boxes in the trays are determined through comparison recognition according to the corresponding medicine box information in the database and the key feature values of the shot medicine box images; after the tray is put on a shelf, the two-dimensional coordinate information of the medicine boxes on the tray is shot by the second camera and is converted into the three-dimensional coordinate information of the medicine boxes through space calculation in combination with the height information of the medicine boxes, and then according to the three-dimensional coordinate of the target medicine box, the line with the shortest goods taking path is found, so that the medicine can be taken. And the accuracy is high.
Owner:SHENZHEN XUEHUI TECH CO LTD

Image auditing method, device and equipment and computer medium

The invention provides an image auditing method, device and equipment and a computer readable medium, and the method comprises the steps: carrying out the character detection of a to-be-audited image through calling a character detection model, cutting the to-be-audited image according to the detected characters, obtaining at least one character image, and carrying out the cutting of the to-be-audited image according to each character image, performing multi-dimensional feature detection on the character image, selecting a corresponding special text recognition model to perform text recognition on the character image according to a feature detection result to obtain a text recognition result, and performing violation judgment according to all the text recognition results corresponding to the to-be-audited image to obtain a violation judgment result. According to the method and the device, the to-be-audited image which is judged to be illegal is processed according to the preset illegal image processing rule, so that an image auditing solution which is high in automation degree, high in accuracy and high in processing speed is provided, the burden of manual auditing can be greatly reduced, and the accuracy and the efficiency of content auditing are improved.
Owner:SHANGHAI BILIBILI TECH CO LTD

Policy interaction relationship identification method and device

The invention discloses a policy interaction relation identification method and device, and the method comprises the steps: inputting a to-be-analyzed policy text into an element extraction model, and obtaining an element set in the policy text; if the number of the elements in the element set is greater than or equal to two, generating an element pair candidate set based on the element set; if the element pair candidate set is not empty, obtaining a relation type and a credibility score corresponding to each element pair through a relation discrimination model; comparing the credibility score with a preset credibility threshold value, and screening out a target element pair from the element pair candidate set; performing time sequence consistency verification on the screened target element pairs; and outputting the target element pair passing the time sequence consistency verification and the corresponding relationship type thereof as the policy interaction relationship of the policy text. According to the method, automatic identification of the policy interaction relationship is realized, and the accuracy and reliability of an identification result are effectively improved.
Owner:HUIZHI TECH (NANJING) CO LTD

Clothing identification method, device, equipment and storage medium

The embodiment of the present invention provides a clothing recognition method, device, equipment and storage medium, the method comprising: extracting key frames at multiple different moments from a video to be recognized; detecting clothing areas and human body areas on the key frames; in response to detecting at least one clothing area and at least one human body area in the key frames, selecting a representative clothing area from the at least one clothing area, the representative clothing area meeting the following conditions: the overlap ratio between the representative clothing area and the human body area reaches a first threshold, the representative clothing area is a target clothing posture, and the clothing exposure ratio reaches a second threshold under the target clothing posture; clothing recognition is performed based on the selected representative clothing area. In this way, the recognition errors that may be caused by unstable clothing postures and incomplete exposure of noise clothing areas are reduced, thereby improving the accuracy of clothing recognition.
Owner:BEIJING IQIYI TECH CO LTD

A training method, application method, apparatus and equipment for a text error correction model

This application discloses a training method, application method, apparatus, and device for a text correction model. The related embodiments can be applied to various scenarios such as artificial intelligence to improve the accuracy of typo recognition. The method of this application includes: outputting a first text probability distribution corresponding to each character in the training sample to be corrected through a text correction model; calculating a correction loss value corresponding to the training sample to be corrected based on the first text probability distribution corresponding to each character; replacing any original character in the training sample to be corrected except for the original typo to construct a noise sample corresponding to the training sample to be corrected; obtaining a second text probability distribution corresponding to each character in the noise sample; calculating a noise loss value corresponding to the training sample to be corrected based on the first and second text probability distributions; and adjusting the parameters of the text correction model based on the correction loss value and the noise loss value to obtain the target text correction model.
Owner:TENCENT TECHNOLOGY (SHENZHEN) CO LTD

An electronic toll collection system based on high-precision recognition

The present invention relates to the technical field of electronic non-stop toll collection, and in particular to an electronic non-stop toll collection system based on high-precision identification, including a data acquisition module, an information processing module, and an identification and deduction module. The present invention greatly improves the accuracy of license plate recognition and the reliability of the system through advanced data acquisition and processing technology. The system collects license plate images through image sensors in the environmental monitoring area, and uses character recognition and judgment units for high-precision recognition to ensure that a high recognition rate can be maintained even under adverse weather conditions. By real-time monitoring of vehicle speed and image clarity, the system can dynamically adjust the heating temperature to prevent image blur and equipment failure. The system automatically adjusts system parameters for interfering weather such as rain and snow, thereby improving the stability of license plate recognition under various climatic conditions. It effectively solves the problem of low accuracy of license plate information recognition under extreme weather conditions due to reliance on light sources and light sensors.
Owner:GUANGZHOU SUIHUA EXPRESSWAY CO LTD

Recognition model training methods, recognition methods, computer equipment, and computer-readable storage media

ActiveCN121561467BReduce collection costsReduce collection workloadElectromagnetic wave reradiation
This application discloses a method for training a recognition model, a recognition method, a computer device, and a computer-readable storage medium. The training method includes the following steps: controlling a single-photon lidar to emit photon pulses towards a target scene and receiving one-dimensional single-photon echoes reflected from the target scene, where the target scene includes at least one object to be recognized; constructing a training sample set based on the one-dimensional single-photon echoes; constructing an initial model and inputting the training sample set into the initial model for iterative training; validating the initial model after each iteration, marking the validated initial model as the recognition model, and outputting it. The application method involves inputting the echo signal to be recognized into the trained recognition model, and obtaining the recognition result based on the processing of the recognition model. The recognition result includes the pose and type of the object in the target scene. Therefore, this application can achieve high-precision recognition of long-distance targets.
Owner:HANGZHOU DIANZI UNIV

Underground pipeline damage identification and positioning method and system based on feature hierarchical model

The present invention discloses a method and system for identifying and locating underground pipeline damage based on a feature hierarchy model, and the method includes: S1, arranging a plurality of Raspberry Pi ultrasonic detectors along the direction of the underground pipeline, and constructing ultrasonic pipe wall damage sample data containing labels; S2, constructing a sample feature hierarchy label learning model to extract features from the ultrasonic pipe wall damage sample data to obtain a label feature vector and a weight coefficient vector through an attention mechanism module; S3, relying on the Raspberry Pi ultrasonic detectors arranged on the underground pipeline wall to collect underground pipeline data at the sampling point, input the sample feature hierarchy label learning model, and output the damage identification result, and use the ultrasonic detector to calculate the location information of the sampling point and store it in association with the damage identification result, and filter and output it. The present invention realizes the output of the sampling point positioning position information and the sampling point damage identification association data, which is convenient for management personnel to efficiently record and maintain work.
Owner:SHENHUA BEIDIAN SHENGLI ENERGY

Hierarchical scanning, automatic identification and processing method based on traffic violation snapshot pictures

The invention relates to a hierarchical scanning, automatic identification and processing method based on traffic violation snapshot pictures, and belongs to the field of image processing and traffic management. The technical method aims at improving the scanning efficiency, accurately recognizing traffic violation information and achieving automatic processing and uploading so as to improve the processing efficiency and accuracy of traffic violation snapshot data. Comprising the steps of hierarchical scanning, automatic identification, data verification and uploading, and exception handling and feedback. According to the method, the stored snapshot pictures are quickly positioned and extracted by presetting the folder hierarchy depth and being compatible with various folder hierarchy structures, and each picture is marked, so that repeated scanning and missing detection are avoided, and the scanning efficiency is remarkably improved. The OCR technology is combined with the deep learning algorithm, so that the recognition accuracy is remarkably improved, and particularly, error recognition can be effectively reduced in a complex environment. After identification is completed, the system verifies the accuracy of an identification result through an automatic verification mechanism, seamless connection and processing of data are completed, manual intervention is reduced, and law enforcement efficiency is improved.
Owner:CHONGQING WEIJING TECHNOLOGY CO LTD

Safety shoe mixed line identification method, device and equipment and storage medium

The invention provides a safety shoe mixed line identification method and device, equipment and a storage medium, and the method comprises the steps: obtaining the identification image information of a plurality of to-be-identified safety shoes, obtaining the corresponding contour information of the safety shoes, and determining a corresponding shooting angle; obtaining a preset safety shoe model library, wherein the preset safety shoe model library comprises a plurality of 3D models corresponding to preset safety shoes and preset safety shoe identifiers; obtaining 3D image information corresponding to each 3D model at the shooting angle; the 3D image information at least comprises a plane view, color information and depth information; respectively carrying out feature extraction on the color information and the depth information to obtain color feature information and depth feature information, and marking the color feature information and the depth feature information in corresponding plane views; and feature matching is carried out on the identification image information and the marked plane view, a preset safety shoe identifier corresponding to the successfully matched plane view is a target safety shoe identifier, the target safety shoe identifier is added to the corresponding identification image information, and mixed line identification of the plurality of safety shoes to be identified is completed.
Owner:FUJIAN NANPING AOXIANG SHOE IND CO LTD

Wireless positioning method and device, medium and product

The invention discloses a wireless positioning method and device, a medium and a product, and relates to the field of wireless positioning.The method comprises the steps that whether a propagation path at the current moment is a sight distance path or not is judged according to obtained real-time monitoring data of a positioning base station; if yes, determining the first arrival path of the sight distance path state and the corresponding first arrival path time delay, and determining the distance between the positioning base station and the terminal and the positioning coordinate of the terminal; if not, determining a first arrival path of the non-line-of-sight path state and a corresponding first arrival path time delay based on local cognitive feedback; judging whether the first arrival path of the non-line-of-sight path state determined based on local cognitive feedback has a problem or not; if not, determining the distance between the positioning base station and the terminal and the positioning coordinate of the terminal; and if yes, determining an estimated first arrival path of the non-line-of-sight path state at the current moment and a corresponding first arrival path time delay estimation value based on cognitive feedback of the cloud, and determining a distance between the positioning base station and the terminal and a positioning coordinate of the terminal. The positioning error can be reduced.
Owner:BEIJING JIAOTONG UNIV

Notebook computer startup verification method and device, notebook computer and storage medium

The invention relates to the technical field of electronic equipment, and provides a notebook computer startup verification method and device, a notebook computer and a storage medium. According to the application, the startup signal of the notebook computer is detected and the ambient light brightness is acquired at the same time, so that the notebook computer can automatically adjust the acquisition brightness of the infrared camera according to the intensity of the ambient light brightness; the self-adaptive adjustment of the acquisition brightness of the infrared camera ensures that clear face images with appropriate contrast can be captured in different illumination environments. The collected face image is subjected to light reflection removal processing, so that the influence of factors such as ambient light and screen light reflection is eliminated or reduced, and the quality of the face image is further improved. According to the method, the notebook computer is subjected to startup verification based on the high-quality target image, so that identification errors caused by the image quality problem are reduced, and the safety of the notebook computer is enhanced. The technical problem of face verification failure caused by dynamic change of ambient light brightness is solved.
Owner:LANCHEN TECH (GUANGDONG) CO LTD

Underground cable ground penetrating radar image automatic identification method based on similarity index

The invention relates to an underground cable ground penetrating radar image automatic identification method based on a similarity index, and the method comprises the steps: carrying out the detection of a plurality of points randomly selected in a detection region through a ground penetrating radar, and obtaining a plurality of ground penetrating radar image data; processing the ground penetrating radar image data to obtain a single-channel waveform under the electromagnetic wave radiation of the ground penetrating radar; the method comprises the following steps: detecting areas where typical media are located, collecting single-channel waveforms of the typical media, and constructing a single-channel waveform library of the typical media; calculating a similarity index between a single-channel waveform obtained by the ground penetrating radar in a region with an unknown medium type and a waveform in a single-channel waveform library of a typical medium; and determining the medium type of the point in the detection area according to the magnitude of the similarity index. According to the method, the medium type of the detection area is determined in combination with the calculation result of the similarity index, cable identification errors possibly caused by human subjective factors are reduced, meanwhile, the calculation process is simple and convenient, the required result can be rapidly obtained, and high effectiveness is achieved.
Owner:STATE GRID HUBEI ELECTRIC POWER CO LTD WUHAN POWER SUPPLY CO