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

137results about How to "Realize precise identification" patented technology

Textile fabric product defect detection method and system

The invention discloses a textile product defect detection method and system, and the method comprises the steps: collecting a textile surface image, and carrying out the preprocessing; constructing a feature extraction model, and extracting multi-modal features of the textile fabric surface defects; analyzing the extracted features, and classifying defect types; positioning defect positions, and marking defect areas and types; and outputting a defect detection report, and optimizing model parameters according to a detection result. Through multi-modal feature extraction and a feedback optimization mechanism, the classification accuracy and the positioning precision of textile fabric defect detection are remarkably improved, and particularly, the method has outstanding performance in detection of complex and low-contrast defects. According to the invention, the detection efficiency is greatly improved, false detection and missing detection are reduced, and an efficient and intelligent solution is provided for quality control of the textile industry.
Owner:SUZHOU HONGYANG TEXTILE DYEING & FINISHING

Slope catastrophe early warning method and system

The invention relates to the field of geological disaster monitoring and early warning, in particular to a slope catastrophe early warning method and system. The method comprises the following steps: establishing a space-time database, storing and managing preprocessed slope data from multiple monitoring sources, and sorting to obtain a slope catastrophe data set; performing feature extraction on the slope catastrophe data set to obtain a key information set reflecting slope catastrophe features; analyzing the key information set to obtain a slope catastrophe value; calculating an early warning threshold value and a real-time displacement prediction value of the side slope based on the side slope data in the space-time database, and fusing the real-time displacement prediction value and the side slope catastrophe value to obtain a fusion value; and evaluating the real-time stable state of each monitored section of the slope according to the obtained fusion value and an early warning threshold value, and judging whether to send out an early warning signal and an alarm signal or not according to an evaluation result. According to the invention, accurate identification of the slope catastrophe value and the potential risk area is realized, and the accuracy and reliability of the early warning system are improved.
Owner:ZHONGJIAO ROAD & BRIDGE (HEBEI) CO LTD

Electromagnetic environment simulation method and system of display module, medium and product

The invention discloses an electromagnetic environment simulation method and system of a display module, a medium and a product, and relates to the field of electromagnetic variable measurement, and the method comprises the steps: receiving a test identification code inputted by a user, extracting a plurality of target scenes corresponding to the test identification code from a preset scene library, and generating a scene change sequence; determining a radio frequency interference source combination and a mechanical vibration parameter corresponding to the target scene; setting test parameters of the test time periods corresponding to the multiple target scenes, and generating a dynamic test sequence; receiving a key test scene marked by a user, and adjusting a test time period and test parameters corresponding to the key test scene to obtain a target test sequence; simulating an electromagnetic environment of the target display module, and recording simulation test parameters of the target display module; the simulation test parameters comprise display brightness, contrast and response time; and determining an electromagnetic compatibility evaluation result based on the target test sequence and the simulation test parameters. According to the invention, the accuracy of an electromagnetic compatibility test result can be improved.
Owner:ALTRON OPTOELECTRONICS (SHENZHEN) CO LTD

Method and device for predicting service life of chemical filter

The invention relates to the technical field of life prediction, and discloses a method and device for predicting the service life of a chemical filter, and the method comprises the steps: collecting the historical multi-dimensional use state data of the chemical filter, and carrying out the standardization preprocessing, and obtaining a time sequence feature data set; carrying out load feature clustering analysis to obtain a filter load mode feature vector; constructing a double-layer performance attenuation prediction model, and outputting a target performance attenuation prediction sequence through physical attenuation coefficient calculation and deep learning network training; performing filter state probability transition calculation to obtain a state transition prediction sequence; performing multi-target joint optimization on the state transition prediction sequence, and calculating to obtain an optimal operation control parameter; and real-time multi-dimensional use state data is collected based on the optimal operation control parameters, parameter optimization is performed on the double-layer performance attenuation prediction model, a service life prediction model is obtained, and the method improves the accuracy and reliability of chemical filter service life prediction.
Owner:SHENZHEN ESKY CLEAROOMS TECH CO LTD

Patient gait abnormity evaluation auxiliary system based on human body posture detection

InactiveCN119132575AAchieve quality improvementAchieve precise labelingMedical automated diagnosisCharacter and pattern recognitionHuman bodyData set
The invention discloses a patient gait abnormity evaluation auxiliary system based on human body posture detection, and relates to the technical field of intelligent medical treatment, and the system comprises the steps: collecting patient gait video data, preprocessing the patient gait video data, and marking patient skeleton key points; based on the patient skeleton data set, calculating position coordinates of patient skeleton key points, and according to the position coordinates of the patient skeleton key points, identifying motion tracks of the patient skeleton key points; on the basis of the patient skeleton key point motion trail, patient gait features are extracted, and a comprehensive gait feature vector is formed; training a gait prediction model by using the patient skeleton data set, inputting the comprehensive gait feature vector into the gait prediction model, and predicting a gait anomaly index of the patient; and evaluating the prediction result, and generating an evaluation report. The gait abnormity index of the patient is predicted by recognizing the motion trail of the skeleton key points of the patient, comprehensive evaluation of the gait state of the patient is achieved, and clinical diagnosis accuracy and efficiency are improved.
Owner:南昌大学第一附属医院

Acute myocardial infarction and atrial fibrillation coexistence risk prediction model construction method and application

The invention provides a prognosis risk prediction model construction method based on coexistence of acute myocardial infarction and atrial fibrillation and application. The method comprises the following steps: S1, acquiring historical patient clinical information of a target patient and death time after coexistence of diseases; s2, preprocessing clinical information of historical patients; s3, using an LASSO method and a ten-fold cross validation method to identify related clinical features related to death time after coexistence of diseases; s4, dividing a training set and a verification set; s5, performing training optimization on the four machine models by using the training set, evaluating the prediction performance of the models by using the verification set, and determining a Glmnet model with the optimal prediction performance; and S6, performing model interpretation on the Glmnet model by using an SHAP value, and determining an optimal prediction factor of which the model prediction contribution ranks the first three. The prognosis risk of coexistence of acute myocardial infarction and atrial fibrillation can be accurately predicted, and reference is provided for clinical diagnosis and treatment of patients with coexistence of acute myocardial infarction and atrial fibrillation.
Owner:WUHAN SHENGBOHUI INFORMATION TECH CO LTD +1

Probe contact impedance real-time monitoring and compensating method and device and storage medium

The invention relates to a probe contact impedance real-time monitoring and compensating method and device and a storage medium. The method comprises the following steps: acquiring a first measurement data set containing a probe contact resistance value and a contact force, and inputting the first measurement data set into a double-layer deep learning neural network for training to obtain a probe contact resistance monitoring model; performing two-parameter fault detection to obtain a contact state classification result; generating a two-dimensional compensation matrix based on a contact state classification result, and performing nonlinear mapping calculation on the second measurement data set to obtain a compensation coefficient; the compensation coefficient, the original probe structure parameters and the original probe operation parameters are input into a graph attention stacking auto-encoder for parameter prediction, and target probe structure parameters and target probe operation parameters are obtained; cloud model calculation and weighting operation are carried out, a target compensation instruction is obtained, and the target compensation instruction is used for adjusting the probe position parameters and the probe pressure parameters. According to the invention, the probe testing precision of the chip is improved.
Owner:东莞市台易电子科技有限公司

Automatic fault detection and repair integrated circuit design method based on neural network

The invention discloses an automatic fault detection and repair integrated circuit design method based on a neural network. The method comprises the following steps: S1, obtaining and preprocessing operation data of an integrated circuit; s2, analyzing the preprocessed data by using a trained neural network model, and dynamically optimizing the model through an adaptive weight pruning algorithm; s3, inputting circuit operation data in real time, automatically identifying fault types and positions, and storing fault information; s4, starting an automatic repair mechanism based on the stored fault information, and generating a repair strategy through a multi-target collaborative optimization algorithm; s5, logic reconstruction is executed according to the repair strategy, a redundant circuit and parameter adjustment are started, and a repair path is optimized through deep reinforcement learning; s6, after repairing is completed, the neural network model is used for detecting and verifying the repairing effect again; and S7, after the repair verification is passed, updating the fault historical database. According to the invention, the automation level of fault detection and repair of the integrated circuit is improved, and the reliability and efficiency of the system are enhanced.
Owner:HANGZHOU SHANGYU ZHIXING TECHNOLOGY CO LTD

Energy management and recovery method and system for cold chain supply chain

The invention relates to the technical field of cold-chain supply chains, and discloses an energy management and recovery method and system for a cold-chain supply chain. The method comprises the following steps: acquiring and preprocessing operating parameters and environmental parameters of the cold chain system to obtain multi-dimensional feature data; performing feature extraction and local density distribution calculation on the multi-dimensional feature data to obtain energy consumption mode clustering data; constructing an equipment operation knowledge graph according to the energy consumption mode clustering data; performing abnormal feature analysis on the equipment operation knowledge graph and the energy consumption mode clustering data through a hybrid neural network model to obtain equipment energy consumption abnormal feature data; generating an initial energy recovery scheme according to the equipment energy consumption abnormal feature data; the initial energy recovery scheme is applied to the cold chain system, the operation efficiency data and the load matching data are collected, online iterative optimization is carried out, the target energy management scheme is obtained, dynamic adjustment and continuous optimization of the energy recovery scheme are achieved, and the energy utilization efficiency of the cold chain system is improved.
Owner:SHENZHEN QIANHAI YUESHI INFORMATION TECH CO LTD

Dangerous rock mass collapse monitoring analysis early warning system and method

The invention relates to the technical field of dangerous rock mass collapse monitoring, and provides a dangerous rock mass collapse monitoring, analysis and early warning system and method.The method comprises the following steps that displacement data and geometrical morphology data of a dangerous rock mass are obtained, format conversion, data cleaning and standardization processing are conducted on the displacement data and the geometrical morphology data, and the displacement data and the geometrical morphology data are stored in a data lake; displacement data and geometrical morphology data are extracted from the data lake, the displacement data and the geometrical morphology data are recognized based on the collapse monitoring model, and the current collapse state of the dangerous rock body is obtained; constructing a collapse analysis and prediction model based on an LSTM model, and predicting the current collapse state through the collapse analysis and prediction model to obtain an estimated change trend of the geometric morphology data; predicting the collapse influence range and the collapse duration of the dangerous rock mass based on the collapse analysis and evaluation model in combination with the current collapse state and the estimated change trend; based on the collapse early-warning model, in combination with the collapse influence range and the collapse duration, evaluating the collapse early-warning level of the dangerous rock body; and starting a corresponding early warning response scheme according to the collapse early warning level of the dangerous rock mass. According to the method, the collapse state of the dangerous rock body is monitored in real time and accurately recognized, the prediction precision of the collapse risk of the dangerous rock body along with the time change trend is improved, and dynamic evaluation and timely response of the collapse early warning level are completed.
Owner:HUBEI GAOTONG SPACE TECH CO LTD

Mining smoke intelligent early warning device, early warning system and early warning method

The invention relates to a mine smoke intelligent early warning device, early warning system and early warning method. The device is characterized in that an aerosol measuring light chamber, a particle high-efficiency filter, a trace gas monitoring chamber and a piezoelectric micropump are sequentially mounted in an inner cavity of an outer shell from front to back; a 650nm laser source, a first detection assembly, a 450nm laser source, a second detection assembly, a 850nm laser source, a third detection assembly and a temperature sensor are respectively arranged in the aerosol measurement light chamber; a piezoelectric quartz crystal detection matrix, a quartz wafer oscillation module and a high-precision frequency measurement module are arranged in the trace gas monitoring chamber; according to the system, a central server is respectively connected with a visual client and a plurality of mining smoke intelligent alarm devices; the method comprises the following steps: monitoring a plurality of monitoring areas by using a plurality of mining smoke intelligent alarm devices; and the central server processes the monitoring data and outputs a monitoring result. The device, the system and the method can accurately monitor smoke in a working place in real time.
Owner:SHENHUA SHENDONG COAL GRP +2

Segmented multi-sensor detection hybrid power transmission line fault rapid positioning and diagnosis system

The invention relates to the field of hybrid power transmission line fault maintenance, in particular to a sectional type multi-sensor detection hybrid power transmission line fault rapid positioning and diagnosing system, which comprises the following modules: a distributed fault positioning module, a fault diagnosis module, a fault diagnosis module and a fault diagnosis module, rapidly positioning a fault point according to the instantaneous traveling wave signal of the line by adopting a traveling wave measurement technology; and the data acquisition and transmission module is used for summarizing the line instantaneous traveling wave signal data acquired by each sensor and transmitting the data to the intelligent data analysis platform. Through the distributed fault positioning module, a fault point can be quickly positioned in an extremely short time, the fault searching time is greatly shortened, the emergency response capability of a power grid is improved, the overall coverage of the whole hybrid power transmission line is ensured through the installation mode of the distributed sensor, and no matter which position the fault occurs, the fault point can be quickly positioned. And fault signals can be captured and processed in time.
Owner:STATE GRID HUBEI ELECTRIC POWER CO LTD WUHAN POWER SUPPLY CO

Method and device for improving GNSS positioning precision, computer equipment and medium

The embodiment of the invention provides a method and a device for improving GNSS (Global Navigation Satellite System) positioning accuracy, computer equipment and a medium, and relates to the technical field of satellite navigation, and the method comprises the following steps: dividing an infrared fisheye picture into a sky region and a shelter region by adopting a sky segmentation algorithm, and generating a sky segmentation map; obtaining a satellite position of a satellite under the earth-centered earth-fixed coordinate system, converting the satellite position from the earth-centered earth-fixed coordinate system to a pixel coordinate system, generating satellite pixel points under the pixel coordinate system, projecting the satellite pixel points into the infrared fisheye image, and generating a satellite coordinate projection drawing; calculating the proportion of the shielding object area in a set range, and judging whether the satellite is a non-line-of-sight satellite or not according to the proportion; and performing weight reduction processing on the observation value through a non-line-of-sight signal equivalent weight matrix, and correcting an output GNSS positioning result. According to the scheme, the observation value of the non-line-of-sight satellite is subjected to weight reduction processing, so that the precision of GNSS positioning is improved.
Owner:Chinese People's Liberation Army Cyberspace Force Information Engineering University

Abnormal region boundary determination method

The invention provides an abnormal region boundary determination method, and the method comprises the steps: carrying out the analysis and processing of a neural signal, obtained by an array electrode, of a target object, and obtaining the boundary point coordinates of a neural signal abnormal region; carrying out segmentation through a tomography image of the target object and a deep learning image segmentation model to obtain boundary point coordinates of an image abnormal region; according to the coordinate of the boundary point of the image abnormal region, optimizing the boundary determination parameter of the neural signal abnormal region, and continuing to optimize the image segmentation model; and comparing a pixel point region with the highest confidence coefficient of the re-optimized image segmentation model with a neuron clustering result, stopping iteration when a comparison result error is smaller than a preset expected threshold value, and determining a boundary of an abnormal region of the target object through image segmentation of the image segmentation model. According to the method, through mutual optimization and iteration of two boundary determination ideas of neural signals and a tomography image + image segmentation model, accurate recognition of the abnormal region boundary can be realized.
Owner:WUHAN NEURACOM TECH DEV CO LTD

Pipeline welding device and method based on visual sensor

The invention discloses a pipeline welding device based on a visual sensor and a welding method, and belongs to the technical field of pipeline welding. The pipeline welding device comprises a welding table, a welding robot is arranged on the welding table, a driving mechanism for driving the welding robot to translate in the length direction of the welding table is arranged on the welding table, and a welding gun assembly is installed on the welding robot; the welding gun assembly comprises a fixed mounting frame, a molten pool monitoring sensor, a welding gun, a welding seam contour sensor and a welding seam positioning sensor. The invention further discloses a pipeline welding method based on the visual sensor. A to-be-welded pipeline can be photographed through a welding seam positioning sensor, then the position of a welding seam is recognized through an AI algorithm model, and a 2D camera is adopted for workpiece recognition and welding seam coarse positioning before welding; the welding seam position is recognized through the AI algorithm model and converted into the position coordinates of the welding robot, the welding robot is guided to conduct welding seam scanning, the positioning precision can be better than 5 mm, and the welding precision is improved.
Owner:CHINA NAT OFFSHORE OIL CORP +1

Fatigue detection method and device based on time sequence self-attention and auxiliary supervision

The embodiment of the invention provides a fatigue detection method and device based on time sequence self-attention and auxiliary supervision. Multi-dimensional features such as the eyelid aspect ratio, the mouth opening degree and the head turning angle of the face of a driver are extracted through face detection and key point positioning. A multi-layer self-attention coding network is innovatively constructed, two branches of trunk feature extraction and auxiliary supervision are included, and a dynamic change rule of a face state is captured by using a time sequence self-attention mechanism. An auxiliary supervision branch is introduced to predict each face parameter state, and a feature learning process is optimized through a multi-task loss function. According to the method, the limitation of traditional single feature analysis is broken through, accurate recognition of the fatigue state of the driver is achieved, and reliable technical support is provided for improving road traffic safety.
Owner:UNIVERSAL UBIQUITOUS TECH CO LTD

Intelligent security and protection monitoring platform based on computer and system thereof

The invention relates to the field of intelligent security and protection monitoring, and particularly discloses an intelligent security and protection monitoring platform and system based on a computer, and the method comprises the steps that an information collection unit collects a security event type, a dangerous event type and real-time monitoring data obtained by a plurality of cameras; the information processing unit screens and acquires abnormal behavior information according to the preprocessed security event type and dangerous event type and marks the abnormal behavior information; performing abnormal behavior information identification on the plurality of cameras; the network construction unit determines a target camera and constructs a target camera network; the security and protection monitoring unit obtains a real-time monitoring data set of all target cameras in the target camera network within the time period; the information analysis unit performs target detection analysis on the real-time monitoring data set acquired by the plurality of target cameras, and judges whether an abnormal target individual exists in a target frame: if yes, acquiring a feature coding sequence of the abnormal target individual under the plurality of different cameras; and if not, continuing security and protection monitoring.
Owner:ZHEJIANG TONGCHUANG INFORMATION SYSTEM CO LTD

Fan acoustic mode decomposition method and system based on multi-measurement Bayesian compressed sensing

The invention discloses a fan acoustic mode decomposition method and system based on multi-measurement Bayesian compressed sensing, and the method comprises the steps: 1, obtaining a multi-measurement Bayesian compressed sensing model according to the highest order and modal monitoring range of an aero-engine fan single-tone noise circumferential dominant acoustic mode; the maximum number of microphones arranged in the circumferential direction of the aero-engine fan casing and the installation position angles of the microphones are determined; 2, a given number of microphone positions are randomly selected from the determined installation position angles, a multi-measurement-vector Bayesian compressed sensing model is constructed according to the modal monitoring range and the selected microphone positions, and the given number is not larger than the maximum microphone number; and step 3, based on the multi-measurement-vector Bayesian compressed sensing model, measuring the circumferential dominant sound mode of the aero-engine fan. The method can automatically adapt to different rotating speed working conditions for acoustic modal measurement, and has better robustness.
Owner:XI AN JIAOTONG UNIV +1

Remote sensing image change detection method and device based on space-time multi-granularity fusion

The invention discloses a time-space multi-granularity fusion remote sensing image change detection method and device, and the method comprises the steps: inputting an obtained remote sensing image data set into a constructed time-space multi-granularity fusion network (STMINet) for training, obtaining a trained STMINet, and storing the trained STMINet; and inputting a remote sensing image to be processed, and performing prediction through the trained STMINet to obtain a dual-time-phase change prediction map.
Owner:ZHEJIANG UNIV OF TECH

Precision detection device for gear processing

The invention relates to a precision detection device for gear processing, which is suitable for detecting the dynamic performance and single tooth precision of a gear, and comprises a base, a rotating bracket, a fixed block, a motor, a reduction gearbox, a test gear, a to-be-tested gear, a main measuring rod, a plurality of measuring rods and the like. High-precision detection of gear dynamic balance, vibration parameters and the tooth outer end and the contour line of a single gear is achieved, the reduction gearbox reduces the rotating speed of the gear and strengthens torque control, and the gear thickness difference design is matched, so that the multiple measuring rods can slide along the tooth outer end of the gear, tooth surface deformation and the contact state are accurately detected, and the device is made of high-carbon low-alloy tool steel. The gear detection device has high strength and wear resistance, adapts to various gear specifications, can output detection data and analysis results in real time, is simple to operate, high in detection precision and high in applicability, and can be widely applied to the fields of gear manufacturing, quality detection and maintenance.
Owner:DONGGUAN KAIMEIQI HARDWARE PRODUCTS CO LTD

Reference marking system, three-dimensional reconstruction method and electronic equipment

The invention provides a reference marking system, a three-dimensional reconstruction method and electronic equipment, and relates to the technical field of three-dimensional reconstruction. The marking assembly comprises at least four non-coplanar markers with different target features, and the target features of the markers have projection stability; the marking assembly has full-view projection specificity, and two-dimensional projection images of the marking assembly are in one-to-one correspondence with three-dimensional space postures of the marking assembly. The reference marking system is suitable for X-ray imaging and natural light imaging, and can realize accurate identification of projection positions of all markers, so that accurate spatial position and attitude information can be provided for three-dimensional reconstruction, high-precision three-dimensional reconstruction can be performed from a single or sparse angle, and the spatial positioning precision and the reconstruction efficiency are improved.
Owner:李尧

Rock mineral content detection system and method for geological exploration

The invention discloses a rock mineral content detection system and a rock mineral content detection method for geological survey, which are characterized in that after image features and spectral features of rocks and ores are extracted, internal generality of the image features and the spectral features of the rocks and ores is mined by using a data processing technology based on deep learning, and the internal generality of the image features and the spectral features of the rocks and ores is used as a mask; according to the method, unique information in rock and ore image features and spectral features is emphasized, and the general information and the unique information of the rock and ore image features and the spectral features are comprehensively utilized through compensation type interactive fusion, so that accurate identification of rock and ore categories is realized. In this way, the problem of information loss in the multi-modal feature fusion process can be effectively solved, and therefore the accuracy of rock and ore classification and the calculation precision of the mineral content are improved.
Owner:山东省地质矿产勘查开发局第一地质大队(山东省第一地质矿产勘查院) +1

Water turbine vibration fault diagnosis method based on time-frequency analysis and residual network

The invention discloses a water turbine vibration fault diagnosis method based on time-frequency analysis and a residual network, and relates to the technical field of fault diagnosis, and the method comprises the steps: collecting and preprocessing the vibration data of a water turbine, obtaining the preprocessed vibration data, and constructing a vibration fault data set; performing model training by using the vibration fault data set, and constructing a vibration fault diagnosis model; and performing fault diagnosis based on the vibration fault diagnosis model to obtain a vibration fault diagnosis result. According to the invention, unified processing of vibration data with different lengths is realized, the problem of inconsistent vibration data lengths in actual engineering is effectively solved, and the utilization rate and representativeness of the data are improved; the convolution operation is performed in combination with batch normalization and an activation function, so that the gradient disappearance problem in deep learning is effectively relieved, and data features can be learned more fully; accurate recognition of no fault, single fault and multiple faults is achieved, the misjudgment rate is reduced, and the credibility of the diagnosis result is improved.
Owner:GUIZHOU WUJIANG HYDROPOWER DEV

Thunderstorm and gale identification method and system based on attention mechanism convolutional neural network

The invention discloses a thunderstorm and gale identification method and system based on an attention mechanism convolutional neural network. The thunderstorm and gale identification method comprises the following steps: 1, generating a sample set according to historical weather radar single station base data and thunderstorm and gale event occurrence actual measurement records, and splitting the sample set into a training set and a verification set; 2, training the recognition model by adopting a training set, verifying the generalization ability of the recognition model by adopting a verification set, and selecting an optimal model parameter according to the model performance; the input of the identification model is a radar reflectivity graph and a radar radial velocity graph of a to-be-identified area at the same time, and the output is a thunderstorm and gale estimation label; and 3, generating a radar reflectivity chart and a radar radial velocity chart according to radar data acquired by a real-time weather radar single station, and inputting the radar reflectivity chart and the radar radial velocity chart into the recognition model to obtain a thunderstorm and gale recognition result. According to the method, for nonlinearity, randomness and burstiness of thunderstorm and gale event space distribution, multi-scale information and position information can be effectively utilized, and the thunderstorm and gale recognition accuracy is improved.
Owner:贵州省气象台

Railway construction protection method and system

The invention discloses a railway construction protection method and system, and the method comprises the steps: firstly, setting hyper-parameters based on a construction operation position; secondly, acquiring video stream data of a construction scene in real time based on an unmanned aerial vehicle; secondly, segmenting a railway in the obtained video stream data by adopting an instance segmentation algorithm, and obtaining pixel point coordinates of an uplink track and pixel point coordinates of a downlink track; then, based on the pixel point coordinates of the uplink track and the pixel point coordinates of the downlink track, a fitting straight line of the uplink track and the downlink track is calculated; then, based on the fitting straight line of the uplink track and the downlink track, an electronic fence is automatically generated; and finally, based on the coordinate position of the constructor and the electronic fence, whether the constructor breaks into the dangerous area is judged. According to the method, the instance segmentation algorithm is adopted to segment the railway, the segmentation effect is good, manual drawing of an electronic fence is not needed, and only the unmanned aerial vehicle needs to be carried to a designated construction site without external cooperation.
Owner:CRSC RESEARCH & DESIGN INSTITUTE GROUP CO LTD

End-to-end cervical cell classification integrated framework

The invention discloses an end-to-end cervical cell classification integrated framework. The system mainly comprises four modules: an enhancement module for expanding and preprocessing an original data set; the fusion module is used for training an improved primary learner on the data set and fusing, outputting and generating a new data set; the purification module is used for purifying the new data set to obtain an integration result of the primary learner and retaining misclassified samples; and the relearning module is used for carrying out relearning by using a secondary learner and carrying out final classification. In addition, a self-adaptive data enhancement method is provided in the training stage of the primary learner and the secondary learner. According to the method, through stage-by-stage fine-grained optimization, the problems that existing cervical cell data sets are unbalanced in category distribution and rare in number are solved, and meanwhile, the classification precision is remarkably improved by emphatically learning easily-confused image features through an integrated framework. Experiments show that the method can effectively assist in cervical cancer screening, the workload of pathologists is greatly reduced, and the method has good application value and huge social benefits.
Owner:HOHAI UNIV

Multi-parameter fusion gas building gas leakage early warning method and device

The invention discloses a multi-parameter fusion gas building gas leakage early warning method and equipment, and the method comprises the steps: continuously collecting the real-time flow and real-time pressure information of a gas meter for measuring the gas consumption of gas equipment in a set time period through a computer, and recording the collection time, the real-time flow and the real-time pressure information of the gas meter according to a time sequence; forming an initial data set; setting a pressure threshold, determining a pressure interval division method, classifying the real-time pressure of the initial data set, forming an initial data subset, continuing to collect the real-time pressure data and supplementing the real-time pressure data to the initial data subset, forming a real-time data subset, and constructing a Poisson random variable model; analyzing the distribution condition of the flow characteristic estimation values in the real-time data subset; and determining the distribution mean values of the flow characteristic estimation values in the real-time data subset, and determining whether the gas supply system leaks or not according to the deviation between the mean values. According to the invention, the accuracy of the early warning system is improved, so that the gas is safer to use.
Owner:NORTH CHINA MUNICIPAL ENG DESIGN & RES INST

CNN-based intelligent water pollution identification system

The invention discloses a CNN-based intelligent water pollution identification system in the field of environmental protection and intelligent monitoring, and the system comprises a data collection module, a data processing module, an interference elimination module, and a pollution identification module which is used for carrying out the accurate identification of pollutants through a convolutional neural network. The judgment logic of the interference elimination module is as follows: if the comparison data of the real-time data and the standard data are consistent, the surface of the water body is pollution-free; when the motion trail presents a feature from bottom to top, animal activity or bubble generation is indicated, and the motion trail is identified as an interference item; and when the motion trail is synchronous with the water surface fluctuation and changes along with the change of the sun position, the motion trail caused by the light and shadow change is identified as the interference item. According to the invention, through a plurality of modules, accurate identification of water pollutants is realized, interference of non-pollution factors is effectively eliminated, and monitoring accuracy and real-time performance are improved.
Owner:CHONGQING TECH & BUSINESS UNIV

ISAR target identification method based on double-branch scattering feature fusion model

The invention provides an ISAR (Inverse Synthetic Aperture Radar) target identification method based on a double-branch scattering feature fusion model. The method comprises the following steps: preprocessing target ISAR data; dividing the current ISAR data into a training data set and a test data set, and training the training data set; constructing a double-branch scattering feature fusion model, and extracting high-dimensional scattering features of an ISAR image domain; inputting the ISAR image and the corresponding ISAR echo data into a double-branch scattering feature fusion model to obtain a corresponding high-dimensional feature tensor, and fusing the high-dimensional feature tensors of double branches to determine a target component segmentation prediction result; and performing model training by using the target component segmentation prediction result and the high-dimensional feature tensor to determine and load a model weight, and obtaining an ISAR target component recognition result by using the test data set. According to the invention, a few-sample semantic segmentation framework based on a double-branch scattering feature fusion model is provided, and accurate identification of a target component under a few-sample condition is realized.
Owner:PLA PEOPLES LIBERATION ARMY OF CHINA STRATEGIC SUPPORT FORCE AEROSPACE ENG UNIV

Method and system for identifying abnormal test data of gyrotron

The invention discloses a gyrotron abnormal test data identification method and system, and the method comprises the steps: 1, constructing a gyrotron test data set, and initializing the parameters of a feature extractor and a Gaussian mixture model; 2, preheating training is conducted, characterization learning is conducted on gyrotron test data through a feature extractor, potential features of the data are obtained, whether preheating training is completed or not is continuously judged, if not, a learning target is continuously optimized, and if yes, the step 3 is executed, and the feature extractor is a deep neural network; 3, after preheating is finished, a Gaussian mixture model is used for conducting distribution modeling on potential characteristics of gyrotron test data; 4, calculating the distribution confidence coefficient of the gyrotron test data by using a Gaussian mixture model, and identifying the gyrotron abnormal test data; and 5, repeating the steps 2-4 until the target training function is converged. According to the method, the accuracy of a gyrotron fault identification task is remarkably improved, abnormal data points in gyrotron test data can be identified, and a powerful guarantee is provided for stable operation of an ECRH system.
Owner:HEFEI INSTITUTE OF PHYSICAL SCIENCE CHINESE ACADEMY OF SCIENCES