38results about How to "Low resolution accuracy" patented technology

The application discloses a data processing method and device, a storage medium and an electronic device. It relates to the technical field of medical treatment. The method comprises the following steps: a tibia three-dimensional model is constructed, and a plurality of target steel nail three-dimensional models are obtained; a plurality of second medical images are obtained, and a plurality of target steel nail first two-dimensional pixel point sets and a tibia second two-dimensional pixel point set are obtained according to the plurality of second medical images; target space conversion parameters of the plurality of target steel nails in the tibia coordinates are obtained by calculation according to the target steel nail first two-dimensional pixel point sets, the tibia second two-dimensional pixel point set, the tibia three-dimensional model and the steel nail three-dimensional models; a guide plate model is obtained according to the target space conversion parameters, and a target space posture of a tibia osteotomy plane is obtained according to the guide plate model. Through the application, the technical problem that the space posture of the tibia osteotomy plane is relatively low in accuracy because the space posture of the tibia osteotomy plane is determined by relying on artificial experience is solved.

The invention relates to a user data processing method and system, a storage medium and electronic equipment, and relates to the technical field of big data process.The method comprises the steps that data cleaning is conducted on original user attribute data of to-be-classified users to obtain target user attribute data, the to-be-classified users are subjected to bucket dividing according to the target user attribute data, and the buckets are obtained; obtaining a user bucket dividing result; determining an attribute association relationship between user entities corresponding to the to-be-classified users included in the same user bucket dividing result, and constructing an entity triple according to the attribute association relationship and the user entities; determining collaborative association edges among the user entities according to the entity triples, and constructing a user association feature map based on the collaborative association edges, the user entities and the target user attribute data; and inputting the user association feature map into a preset user classification model to obtain a user classification result, and determining the user category of the to-be-classified user according to the user classification result. The accuracy of the user classification result is improved.

The invention discloses a data processing method and device, a storage medium and electronic equipment. The method comprises the following steps: constructing a tibia three-dimensional model, and obtaining a steel nail three-dimensional model of a plurality of target steel nails; obtaining a plurality of orthogonal second medical images, and obtaining a first two-dimensional pixel point set of the plurality of target steel nails and a second two-dimensional pixel point set of the tibia according to the plurality of second medical images; calculating according to the first two-dimensional pixel point set, the second two-dimensional pixel point set, the tibia three-dimensional model and the steel nail three-dimensional model to obtain target space conversion parameters of the multiple target steel nails under the tibia coordinates; and obtaining a guide plate model according to the target space conversion parameters, and obtaining a target space attitude of the tibia osteotomy plane according to the guide plate model. By means of the method and device, the technical problem that the accuracy of the space posture of the tibia osteotomy plane is low due to the fact that the space posture of the tibia osteotomy plane is determined through related technologies depending on artificial experience is solved.

The application discloses a hard disk fault prediction method and device, a storage medium and an electronic equipment, and relates to the technical field of computer storage. The method comprises the following steps: acquiring historical state data of a plurality of storage hard disks recorded in a server data center, wherein the historical state data comprises hard disk fault data and hard disk non-fault data; performing trajectory guidance processing on the hard disk fault data by using a target diffusion generation model to obtain virtual fault data; determining a training data set based on the virtual fault data and the historical state data, training a target prediction model by using the training data set, and obtaining a hard disk prediction model; and in the case of determining to predict a target storage hard disk, inputting real-time hard disk state data of the target storage hard disk in a previous period into the hard disk prediction model, and predicting fault information of the target storage hard disk in a next period by using the hard disk prediction model. The problem of low accuracy of hard disk fault prediction is solved.

This invention discloses a machine learning-based method and apparatus for automatic ETL task configuration. The method involves collecting historical data and corresponding configuration results, training the ETL configuration model using machine learning, and extracting the mapping relationships of task keyword groups from the historical data as a task rule base. The task rule base is then used to match target ETL tasks, obtaining the matching degree of different configuration results. When the matching degree of all different configuration results is less than or equal to a preset threshold, the ETL configuration model is used to match configuration results for the target ETL task. This approach improves the accuracy of matching results and reduces computational cost by implementing most of the matching tasks using a simple keyword-based task rule base, while still providing accurate matching results.

The invention relates to the technical field of computer software, and discloses a code component generation method and device, equipment, a medium and a product, and the method comprises the steps: obtaining a target text and data assets matched with the target text; and when at least one target data asset is associated with the associated data asset, determining a business priority corresponding to the target data asset according to the source information corresponding to the target data asset, and determining the business priority corresponding to the target data asset according to the association relationship between the associated data asset and the target data asset and the business priority corresponding to the target data asset. Determining an associated business priority corresponding to the associated data assets; sorting the target data assets and the associated data assets according to the service priorities and the associated service priorities to obtain a sorting result; according to the sorting result, the target data assets and the associated data assets are input into a code component generation model, code components corresponding to the target text are obtained, and the problem that the correct rate of the code components generated by the model in visual programming is low is solved.

The invention provides a question and answer method, device and equipment suitable for a traditional Chinese medicine compound molecular mechanism scene, a medium and a product. The method comprises the following steps: acquiring problems in a traditional Chinese medicine compound molecular mechanism scene; forming first prompt information based on the question and the first prompt template, and inputting the first prompt information into a trained question and answer model to obtain answer information of the question; wherein the first prompt template comprises a multi-step thinking normal form, and the multi-step thinking normal form comprises a plurality of processing steps; the question and answer model generates a plurality of sub-tasks corresponding to the question based on a plurality of processing steps in the multi-step thinking normal form, retrieval is performed in the domain knowledge graph of the traditional Chinese medicine compound molecular mechanism scene based on the plurality of sub-tasks, recall knowledge is obtained, and answer information of the question is generated based on the recall knowledge. The problem that a general question and answer model is low in accuracy in the aspect of traditional Chinese medicine mechanism analysis is solved, and application of the question and answer model in the field of traditional Chinese medicine compound molecular mechanism scene perpendicularity is achieved.

The invention relates to a high gamma oil layer identification method, and belongs to the technical field of oil and gas field exploration and development. The high-gamma oil layer identification method comprises the following steps of determining a high-gamma sandstone section of a typical well drilling stratum and a reservoir type of the high-gamma sandstone section based on typical well data in a research area; then determining an evaluation index for representing the oiliness of the high-gamma sandstone section based on the gas logging response characteristics of the high-gamma sandstone section; and finally, performing rendezvous analysis on the basis of the corresponding relationship between the evaluation indexes of the high-gamma sandstone section and the reservoir types, and constructing an identification standard of the high-gamma oil reservoir in the research area. Experimental results show that when the recognition method is adopted to recognize the high-gamma oil reservoir, the accuracy is high, and the problem that the oil reservoir of the high-gamma oil reservoir cannot be effectively and accurately recognized can be solved.

The application relates to a surgical robot and a control method of the surgical robot, the surgical robot comprising a robot master system, a robot slave system and a scanning device, the robot master system being in communication connection with the robot slave system and the scanning device respectively; the scanning device acquires a real-time scanning image of a scanning object; the robot master system controls the robot slave system to move according to a motion path according to the real-time scanning image and the motion path, wherein the motion path is obtained after being planned according to the position of a region of interest in the real-time scanning image; the problem that the field of view of a lens of a thoracoscope or a laparoscope used for observation is limited in the related art, resulting in reduced accuracy in a surgical process, is solved; and the application provides a larger surgical field of view through the real-time scanning image, and the safety and accuracy of the surgical process are improved.

The present application relates to the technical field of environmental characteristic measurement, and particularly relates to a kind of sea-air environment dual-band infrared radiation characteristic measurement method and device.The method comprises: the portable measuring device is built on the shipborne platform;Control two-dimensional rotary table to rotate according to N type variable speed scanning path, and synchronously control mid-long wave thermal imager to carry out continuous image acquisition, while by photoelectric inertial navigation real-time record attitude information;Based on the time axis of photoelectric inertial navigation, each frame infrared image is aligned with the corresponding accurate azimuth and pitch angle Data, and the image gray value is converted into radiation brightness;Finally, the results are summarized.The device includes tripod, two-dimensional rotary table, adapter plate and the mid-long wave thermal imager and photoelectric inertial navigation fixed thereon.The present application synchronizes and aligns the data of photoelectric inertial navigation and thermal imager, effectively compensates the disturbance of ship body posture, solves the problem of low accuracy of infrared radiation measurement data on dynamic platform, and simultaneously has the advantages of fast response and portability.

The disclosure provides a traffic signal light switching point prediction method and device, electronic equipment and storage medium, relates to the technical field of data processing, specifically relates to big data, intelligent transportation and cloud computing and the like, and can be applied to the technical field of navigation, Internet of Vehicles, intelligent cockpit and automatic driving, and comprises the following steps: acquiring vehicle trajectory data of a target intersection; calculating switching point prediction correlation information of a target traffic signal light in the target intersection according to the vehicle trajectory data; wherein the target traffic signal light comprises a straight signal light and / or a non-straight signal light; the switching point prediction correlation information comprises signal light switching rule information, signal light phase difference information and signal light real-time switching point information; and predicting switching point information of the target traffic signal light according to the switching point prediction correlation information of the target traffic signal light. The embodiment of the disclosure can improve the accuracy and efficiency of traffic signal light switching point prediction.

The application relates to an image display method and device, an AR display system and an electronic device, and is applied to an AR display system, the AR display system comprising a diffractive optical waveguide. The image display method comprises the following steps: acquiring an image to be displayed; inputting the image to be displayed into a pre-trained image conversion model to output a target compensation image; wherein the image conversion model is generated according to at least a projection inverse function; the projection inverse function is used to indicate an image mapping relationship between the image to be displayed and the target compensation image, and the image mapping relationship is at least associated with the diffractive optical waveguide; and generating an image display result based on the target compensation image. Through the application, the problem of low accuracy of image display is solved, and a display optimization method of a high-precision AR display system is realized.

This application relates to an image detection method, apparatus, system, and electronic device. The image detection method includes: acquiring a white light image and a NBI image of an image to be detected; performing feature extraction processing on the white light image and the NBI image respectively to obtain color depth features of the white light image and texture depth features of the NBI image; performing depth feature fusion processing based on the color depth features and the texture depth features to obtain fused depth features; performing image quality detection processing based on the fused depth features; and generating a target image detection result for the image to be detected. This application solves the problem of low accuracy in image detection and achieves an accurate and efficient image detection method.

This application provides an image positioning correction method, apparatus, storage medium, and electronic device. The method includes: acquiring an image to be positioned; selecting a calibration image corresponding to the image to be positioned from a preset set of calibration images to obtain a target calibration image, and identifying a set of calibration points in the image to be positioned; updating the homography matrix based on the set of calibration points in the image to be positioned and the set of calibration points in the target calibration image; correcting the pixel coordinates of each pixel in the image to be positioned according to the updated homography matrix; and positioning the image to be positioned according to the corrected pixel coordinates of each pixel to obtain the geographic coordinates corresponding to the image to be positioned. This application solves the problem of poor positioning accuracy by updating the homography matrix, thereby reducing the error in geographic coordinate mapping.

The invention relates to a carsickness recognition method, an electronic device and a vehicle, and the method comprises the steps: obtaining the environment features of the vehicle and the visual features of a target passenger in the vehicle; the vehicle environment characteristics comprise vehicle state parameters and environment parameters associated with inducing carsickness of passengers in the vehicle, and the visual characteristics are behavior performance characteristics related to the carsickness state; constructing a main feature vector according to the visual features; according to the visual features and the vehicle environment features, auxiliary feature vectors are constructed; based on the correlation degree between the main feature vector and the auxiliary feature vectors, performing dynamic weighted fusion on each auxiliary feature vector to generate a multi-modal feature vector; and carsickness identification processing is carried out on the multi-modal feature vector, and a carsickness identification result for the target passenger is generated. According to the invention, the problem of low carsickness identification accuracy is solved.

This invention relates to the field of data analysis technology, and discloses a method, device, electronic device, and storage medium for analyzing human squatting behavior on a toilet. The method includes: performing a Fast Fourier Transform on millimeter-wave radar echo signals to obtain a range-Doppler image; extracting the posture point cloud features and range-Doppler features of the human target in the range-Doppler image; performing height analysis on the human target based on the posture point cloud features to obtain sitting height thresholds and standing height thresholds, and identifying candidate sitting and standing behaviors; and performing probability discrimination of squatting behavior on the human target based on the candidate sitting and standing behaviors, posture point cloud features, and range-Doppler features to obtain the squatting behavior recognition result. By implementing signal processing, feature extraction, height analysis, and squatting behavior discrimination of millimeter-wave radar signals, this invention can improve the accuracy of human squatting behavior analysis while ensuring the privacy of human behavior analysis.

The invention discloses an in-vitro simulation device and method for ultrasonic mechanical damage. The in-vitro simulation device comprises an ultrasonic transmitting unit, a reaction container, a control unit, a bionic sample and a liquid circulating unit, wherein the bionic sample and the liquid circulating unit are arranged in the reaction container; the bionic sample comprises a plurality of simulated focuses; the control unit is in communication connection with the liquid circulation unit and is used for controlling the working state of the liquid circulation unit according to the obtained preset flow rate so as to control the flow rate of liquid in the reaction container; the control unit is also in communication connection with the ultrasonic transmitting unit and is used for controlling the ultrasonic transmitting unit to transmit ultrasonic waves to a plurality of simulated focuses according to a preset damage sequence and circulating for multiple times; and the control unit is also used for acquiring damage conditions of the plurality of simulated focuses in the ultrasonic action process and adjusting working parameters of the ultrasonic transmitting unit according to the damage conditions. According to the invention, the accuracy of the finally adjusted working parameters of the ultrasonic transmitting unit can be improved.

The invention relates to the technical field of image processing, in particular to an automatic annotation data collecting system which comprises an adjusting module, a recognition module, an annotation module, a processing module and a control module, when images are collected, the recognition module collects the images and character data, the annotation module annotates the images and character data to obtain annotation data, and the processing module processes the annotation data to obtain annotation data; the processing module is used for acquiring a next image, transmitting the annotation data to the processing module, judging the annotation data, acquiring the next image if the annotation data meets a preset condition, and transmitting the judgment data to the control module if the annotation data does not meet the preset condition; and after the adjustment module is controlled to adjust the acquisition angle and the acquisition brightness of the identification module, the identification module performs acquisition again until the annotation data meets the preset condition, so that the problem of low accuracy of an existing image processing mode is solved.

The invention provides an accelerated test method for simulating the corrosion condition of a part in a marine atmospheric environment. The accelerated test method comprises the following steps: putting a test part in a simulation environment for a periodic corrosion test; the periodic corrosion test comprises a damp-heat test, a drying test and a salt spray test, the damp-heat test and the salt spray test are alternately carried out, and the drying test is arranged after each damp-heat test and each salt spray test. Therefore, a periodic corrosion test with alternate wetting and drying is introduced, and a plurality of stages such as wetting, electrochemical corrosion, drying crystallization and the like in the natural corrosion process are simulated more truly. A humid heat test and a salt spray test simulate a corrosion environment with high humidity and salt deposition, and a key drying test promotes conversion of corrosion products and diffusion of oxygen, which is more consistent with a formation process of a corrosion layer in a natural environment; therefore, the corrosion mechanism generated by the accelerated test and the microstructure of the corrosion product have better consistency with the corrosion condition in the real natural environment, and the accuracy of the test result is improved.

The invention discloses a shooting action mode recognition and prediction method and system, and belongs to the technical field of mode recognition and prediction. The method specifically comprises the steps that a shooting action time sequence signal set and working condition information are collected, time alignment and amplitude normalization are carried out, and an aligned shooting action time sequence signal set and a working condition label set are obtained; based on the aligned shooting action time sequence signal set and the working condition label set, spatial and temporal distribution alignment is carried out through optimal transmission, a multi-layer multi-relation hypergraph is constructed, hyperbolic manifold embedding is carried out, and a hyperbolic hypergraph representation sequence and weight are obtained; constructing a conditional energy function based on the hyperbolic hypergraph representation sequence and the weight, performing optimization calculation through a micro-optimization layer to obtain a stage sequence, a mode label and a stage transfer parameter of the shooting action, constructing a multi-time-interval stage hazard rate matrix, and obtaining a mode recognition prediction result; the accuracy of shooting action mode recognition prediction in a new scene is improved, and the stability of shooting action mode recognition prediction is enhanced.

The application discloses a multi-modal residual value evaluation method and system for mobile terminal pre-set-off, and belongs to the field of electric digital processing, comprising the following steps: migratable feature acquisition, migratable feature reliability evaluation, residual value evaluation result verification and mobile terminal pre-set-off. The application analyzes migratable features by inputting static data into a migration learning model, synchronously evaluates the reliability of the migratable features, determines whether to take reliability adjustment measures based on the evaluation results, then performs residual value evaluation to output residual value analysis results for credibility verification, pre-sets off the mobile terminal to be evaluated based on the residual value analysis results after credibility verification, and determines whether to end the lease period based on the reservation mechanism triggered by the mobile terminal to be evaluated after the preset lease period, thereby improving the accuracy of pre-set-off data and solving the problem of reduced accuracy of pre-set-off results due to reduced reliability of residual value evaluation results in the prior art.

The invention relates to an abnormal account identification method and a storage medium, and the method comprises the steps: obtaining the multi-dimensional transaction feature data of a plurality of to-be-identified accounts on a block chain; according to the position data of each to-be-identified account, calculating an HHI index, carrying out feature fusion on the multi-dimensional transaction feature data and the HHI index, and constructing a feature vector matrix corresponding to the to-be-identified account; clustering each account to be identified based on the feature vector matrix to generate a cluster; and determining an abnormal account cluster from the plurality of accounts to be identified based on the clustering cluster. According to the invention, the problem of low accuracy of abnormal account identification is solved.

The application provides a driver state judgment method and device and a driver monitoring system. The method obtains a true value and a predicted value of a regression task in a driver monitoring task. Then, a first continuous probability distribution function is determined according to the true value, a second continuous probability distribution function is determined according to the predicted value, a loss function corresponding to the regression task is obtained based on a loss function corresponding to a classification task and the above functions, a trained neural network model is obtained based on the loss function, the neural network model inputs driver data and outputs a predicted value and a confidence of the driver. Thus, the application can determine the prediction result output by the neural network by using the confidence output by the neural network, improve the accuracy of the driver state judgment based on the prediction result output by the neural network, identify the behavior of the driver that is not conducive to the safe driving of the vehicle in time, and greatly reduce the probability of accidents of the vehicle.

The invention provides a converter dynamic control method and system.The method comprises the steps that an oxygen blowing quantity prediction model is established through a gradient lifting algorithm, then smelting parameters are input into the oxygen blowing quantity prediction model, and the reference oxygen blowing quantity is calculated through the oxygen blowing quantity prediction model; and the coolant adding amount is calculated based on a temperature balance equation, the oxygen blowing amount is calculated according to the coolant adding amount, and finally a control result is generated through the coolant adding amount and the oxygen blowing amount. According to the method, an oxygen blowing amount prediction model is established, calculation is carried out based on a temperature balance equation, intelligent determination of the oxygen blowing amount and the coolant adding amount is achieved, and the problem that the accuracy of oxygen blowing amount calculation is low is solved.

This application discloses a method, system, device, and storage medium for predicting the immunogenicity of neoantigens. The method includes: acquiring tumor mutation data and synthesizing candidate peptides; constructing a training set with HLA typing and immune tags; extracting and fusing multi-omics feature vectors covering dimensions such as binding affinity, complex stability, presentation probability, expression level, mutation frequency, and clonal abundance; training the model using random balanced forest and XGBoost ensemble learning to obtain an immunogenicity prediction model; outputting immunogenicity probability scores for the peptides to be predicted and prioritizing them accordingly to generate a neoantigen list. This application significantly improves the accuracy and screening efficiency of neoantigen immunogenicity prediction by fusing multi-omics features and integrating the immunogenicity prediction model.

This application provides a method and system for predicting peak vibration of open-pit bench blasting, relating to the field of engineering blasting safety technology. The method, after acquiring monitoring data, constructs a multi-parameter formula for predicting the maximum peak velocity of blasting vibration through dimensional analysis. Then, based on the velocity prediction formula, Bayesian theory is used to obtain the optimal prediction model for peak blasting vibration, which is then used to predict the peak blasting vibration. This method introduces Bayesian theory into the field of blasting vibration prediction, achieving a correction of the blasting vibration prediction model that balances data fitting and anti-interference capabilities from both physical and statistical perspectives. This method improves the accuracy of peak vibration prediction, provides a reliable basis for controlling blasting vibration, and solves the problem of low accuracy in peak vibration prediction.

The application discloses a kind of entity relation extraction methods based on BERT and entity position information belonging to the field of natural language processing, which comprises the following steps: step 1: obtaining entity relation extraction text dataset containing entity position mark information, divided into training set, validation set and test set;Step 2: considering the influence of entity position on sentence text features, an entity relation extraction composite model combining BERT, BiLSTM and attention mechanism is constructed;Step 3: the training set in the data set determined in step 1 is used to train the model, and the validation set is used to determine the stop round of training, and the result is used for test set evaluation;Step 4: the model is evaluated using the test set, and the rationality of the model is analyzed.The application realizes more accurate and efficient mining effect, provides a novel way to mine the potential relationship between entities, and brings good versatility and practical value.

The invention discloses a material verification method and device, a storage medium and electronic equipment, and relates to the technical field of smart home, and the material verification method comprises the steps: obtaining the identification information of a target material collected by a radio frequency reader-writer, and the image information of the target material collected by a visual sensor, the first weight information of the target material is collected by a weight sensor; verifying the target material according to the identification information, the image information and the first weight information; under the condition that the target material is successfully verified, a first starting instruction is sent to a target baffle, so that the target baffle is started according to the first starting instruction, and the target material is moved to the first position, and by adopting the technical scheme, the problem that in the related technology, the accuracy rate of material verification is low is solved.

The invention discloses an angle detection method, device and equipment of a sowing mechanism and a storage medium, and relates to the technical field of unmanned aerial vehicles. The sowing mechanism comprises a driving device and a sowing disc, the driving device comprises a motor and a speed reducer, and the motor is connected with the speed reducer and drives the speed reducer to rotate; an output shaft of the speed reducer is connected with the sowing disc, the sowing disc comprises a shell and a throwing disc, a rotor of the motor is connected with an input shaft of the speed reducer, an output shaft of the speed reducer is connected with the throwing disc, the throwing disc is arranged in the shell, the shell is provided with a Hall sensor, the throwing disc is provided with a magnet, and the rotor of the motor is provided with a position sensor. The method comprises the following steps: acquiring an origin angle of an output shaft, wherein the origin angle is determined based on a trigger signal collected by a Hall sensor; obtaining a current rotor angle collected by a position sensor; and determining a current output shaft angle according to the current rotor angle and the original point angle. According to the scheme, the problem that the accuracy of controlling the throwing disc to rotate by the motor is reduced due to poor signal transmission quality in related technologies is solved.