54results about How to "High-precision identification" patented technology

The invention relates to the technical field of image processing, and discloses a detection method, system and equipment for intelligent welding of pipe pile splicing and a medium. The method comprises the steps that noise reduction is conducted on a target weld seam image through a filtering algorithm, registration and splicing are conducted on the image through feature matching, and a weld seam unfolding image is obtained; segmenting the weld joint expanded image by using a deep learning network, and extracting a weld joint center line and surface features to obtain weld joint feature data; defect detection and recognition are conducted on the weld joint feature data, hierarchical calculation is conducted through a rule set, and weld joint quality data are obtained; parameter mapping is conducted according to the weld joint quality data, a welding parameter instruction is generated, welding process variables are adjusted, and target welding parameters are obtained; and the welding seam quality data and the target welding parameters are stored, and a quality report is generated through data analysis. The detection accuracy of intelligent welding for pipe pile splicing is improved.

The invention discloses a smart home voice interaction control method and system, and relates to the technical field of smart home. The problems that in the prior art, the intention of a user cannot be accurately understood through non-voice information such as emotion and expression of the user, a control strategy is solidified and lacks individuation, and user requirements and emotion cannot be met in real time are solved. By fusing voice and non-voice information, accurate recognition of user intentions is enhanced, recognition accuracy of voice instructions is improved, emotional states and requirements of users are understood more comprehensively, particularly in a noisy environment or a complex situation, recognition and response effects of the system are obviously improved, and user experience is improved. By utilizing the emotion analysis and context understanding technology of the user, the interaction mode is dynamically adjusted according to the emotion state of the user, the emotion demands of different users can be flexibly adapted, more humanized services are provided, the control strategy is continuously optimized according to the feedback of the user, the user satisfaction is improved, and the user experience is improved. Therefore, the intelligent degree and the user experience of the smart home system are improved.

The invention provides an image segmentation method and device combining feature difference recognition and detail enhancement, and the method comprises the following steps: obtaining a to-be-segmented image to construct a data set, and carrying out the marking and preprocessing of the data set; constructing an image segmentation model, wherein the image segmentation model comprises an encoder network, a jump connection network, a decoder network and an output network; the encoder network uses a pre-trained CNN (Convolutional Neural Network) and Transform hybrid network to carry out feature extraction; the jump connection network comprises a channel space interaction module for fusing encoder features and decoder features; the decoder network comprises a convolution-channel attention module for extracting key features and a convolution-up-sampling module for up-sampling; the output network comprises a multi-level feature fusion module for fusing the multi-scale features of the decoder; training the image segmentation model to obtain a trained image segmentation model; and segmenting the image data by adopting the trained image segmentation model.

The invention provides a distribution line fault waveform denoising analysis method based on an improved wavelet threshold function and computer equipment, and relates to the technical field of power system fault detection and signal processing. The method comprises the following steps: S1, obtaining a fault waveform of a distribution line through actual measurement or analogue simulation, and converting the fault waveform into a fault signal; s2, through denoising effect comparison, selecting an applicable wavelet basis function and a decomposition layer number, and carrying out wavelet transformation to decompose a fault signal; s3, carrying out wavelet decomposition coefficient threshold processing on the fault signal by adopting a self-adaptive adjustment mechanism of a threshold and an improved threshold function; and S4, performing discrete wavelet inverse transformation on the wavelet decomposition coefficient subjected to threshold processing to reconstruct a signal so as to obtain a denoised fault signal. According to the method, through the improved threshold function and the multi-scale analysis technology, the fault waveform denoising effect can be remarkably improved, the accuracy and the real-time performance of fault detection are ensured, and therefore the reliability and the safety of a distribution line are improved.

The invention discloses a medical image segmentation method based on an edge-guided attention mechanism, and the method carries out the segmentation of a medical image through a multi-scale edge-guided attention image segmentation model. The multi-scale edge-guided attention image segmentation model comprises an encoder, a decoder, an edge-guided attention module and a multi-scale feature map fusion module; by introducing an edge-guided attention mechanism into a multi-scale edge-guided attention image segmentation model, compared with an existing method, the method has the advantages that more edge details can be reserved when an image with a small target and a blurred boundary is segmented; the effectiveness of accurate recognition and boundary form segmentation of colorectal cancer osteoblast metastasis tumor by a multi-scale edge guide attention image segmentation model is improved. Meanwhile, the problems that in the colorectal cancer osteogenesis metastasis tumor segmentation process, correlation between the features is not fully considered and important features are not effectively screened in an existing method are solved, and therefore the accuracy of image segmentation is improved.

The invention discloses a preparation method and application of a concave-structure SERS (Surface Enhanced Raman Scattering) substrate for prostatic cancer exosome detection, and relates to the field of Raman spectrum detection. The invention aims to solve the problems that the quantity of electromagnetic field hot spots of the currently reported SERS substrate with the concave structure is small, the SERS substrate does not have good uniformity, the size is not matched with EVs, the detected EVs signal is relatively weak, and the parting detection requirement cannot be met. The preparation method of the SERS substrate with the concave structure comprises the following steps: 1, preparing a PDMS concave trap structure array; 2, preparing a concave well array of which the surface is covered with densely arranged Au NPs; according to the invention, Au nanoparticles are grown in situ in a PDMS concave structure, a concave trap array of which the size is matched with that of EVs and the surface is covered with densely arranged Au NPs is constructed, high-sensitivity detection (lower detection limit of 105 particles / mL) and high-accuracy typing of prostatic cancer EVs are realized, and the accuracy rate can reach 98% or more.

The invention discloses an agricultural precision planting pre-training and fine-tuning training method based on an AI large model. Based on a deep learning technology, key information is obtained from large-scale agricultural data, an optimal planting mode is predicted, and the crop yield and quality are improved. Through analysis of historical data and real-time monitoring, peasants are predicted and reminded about possible plant diseases and insect pests, so that measures are taken in advance. For different regions and different environmental conditions, the model can adapt to specific planting requirements through fine adjustment, and the universality and accuracy of the model are improved. The prediction of soil, weather and plant growth modes provides intelligent decision support for agricultural production, helps farmers to optimize a planting scheme, improve the yield and reduce resource waste, thereby improving the scientization and production benefits of agriculture. The model can help farmers to scientifically plant, improve crop yield and quality, save resources, reduce environmental burden and promote agricultural sustainable development.

The invention discloses a sewage treatment fault diagnosis method and system based on data analysis, and the method comprises the steps: continuously optimizing a diagnosis strategy in the input and training process of a large amount of data through introducing a fault diagnosis model, and improving the accurate recognition capability of a fault under a complex working condition; therefore, the system not only has strong learning ability, but also can perform self-optimization through a reward mechanism, thereby reducing the occurrence rate of false report and missing report. An adaptive weight adjustment mechanism is introduced, so that the model can automatically adjust weight distribution and increase response to a new fault when the system has a novel fault or few faults; by dynamically adjusting the weight vector, the diagnosis model of the invention has strong self-learning ability and can continuously adapt to new fault modes and working condition changes, and the weight adjustment mechanism greatly improves the generalization ability and adaptability of the model, and especially has significant advantages in the face of fault scenes with strong uncertainty.

The invention relates to the technical field of deep learning, and discloses a potential hot content identification method based on a mainstream value droop domain knowledge base, which comprises the following steps: constructing the mainstream value droop domain knowledge base, carrying out topic modeling through a hidden Dirichlet distribution model and extracting potential topic words, an event triple is extracted through a RoBERTa-CRF (Conditional Random Field) model, and an event atlas is constructed; nodes related to the hot spots are retrieved from the event atlas through an atlas retrieval engine, sub-graphs formed by the related nodes are converted into vector representation through a graph embedding technology, and a hot spot mode is formed; and by calculating Jaccard similarity, matching content features similar to the hotspot mode in the fusion graph representation network, and identifying potential hotspot content. According to the method, the forward enhanced information content is accurately identified and screened through thematic modeling, event extraction, atlas retrieval, similarity calculation and other technologies, and the real-time monitoring and prediction capability on the social hotspot trend can be enhanced.

The invention provides an accurate target identification and path planning method and system based on point cloud segmentation, and relates to the technical field of robot navigation and path planning, and the method comprises the steps: firstly, collecting three-dimensional point cloud data of a target region, and setting related parameters; then, segmenting the point cloud by using a region growing algorithm to obtain a plurality of target regions; thirdly, extracting target features based on a segmentation result, and performing target recognition by using a support vector machine (SVM); then, calculating an optimal path from the starting point to the target point through an improved dynamic weighting A * algorithm; and finally, introducing time sequence information in the path planning process to realize real-time path optimization. Through the technical scheme of the invention, the accuracy of target identification is improved, the efficiency of path planning is optimized, the robot can better navigate in a complex environment, and the method has important application value.

The invention discloses a method for calculating the wave load borne by a pile based on machine vision, and the method comprises the steps: collecting and obtaining a water surface image sequence containing the pile, carrying out the image processing of the water surface image sequence, obtaining a flow field image, carrying out the post-processing of the flow field image, and obtaining an optimized flow field image, the flow field is comprehensively monitored through machine vision, and accurate recognition of the flow field where the pile is located is improved; calibration of a preset image scaling is carried out on the optimized flow field image, pixel coordinates in the optimized flow field image are converted into actual coordinates, pile periphery data under a corresponding flow field are obtained based on a particle image velocimetry (PIV) algorithm, and a velocity vector field of fluid particles can be obtained through the particle image velocimetry algorithm (PIV); the flow velocity distribution, vortex and other fluid mechanics parameters of the whole water area are further deduced, pile periphery data under the corresponding flow field can be obtained, then the wave load borne by the pile is obtained based on the Morison equation, and the defect that according to an existing wave load method, due to the fact that the working environment of underwater ocean engineering structures is complex and severe or the safety requirement is high, the wave load is poor is overcome. The problems of low wave load accuracy and high investment cost are solved, and the accuracy and efficiency of the wave load borne by the pile are greatly improved.

The invention discloses a gas detection method and system based on multi-modal data, and a terminal. The method comprises the following steps: collecting VOCs gas data and corresponding VOCs text data; performing data preprocessing on the VOCs gas data, outputting standardized VOCs gas data, and extracting gas characteristic values from the standardized VOCs gas data; text features are extracted according to the VOCs text data, multi-modal data fusion is carried out according to the gas feature values and the text features, and fusion feature vectors are generated; and constructing a gas recognition model according to the fusion feature vector, inputting the to-be-detected gas data into the gas recognition model, and outputting a detection result of the to-be-detected gas data. On the basis of multi-modal data fusion and deep learning technologies, by integrating sensor data and applying a data processing algorithm, high-precision recognition of a complex gas mixture is achieved, gas information easy to understand can be provided in real time, and higher practicability and application value are provided.

The invention discloses a building fire-fighting monitoring system based on the Internet of Things. The building fire-fighting monitoring system comprises a data acquisition module, a building fire-fighting evaluation model establishment module and a building fire-fighting monitoring module. The invention belongs to the field of data processing, and particularly relates to a building fire-fighting monitoring system based on the Internet of Things, and the scheme is based on a multi-scale time sequence feature extraction and self-adaptive time step scaling mechanism, and achieves the synchronous detection of an instantaneous fire signal and periodic equipment abnormality; based on time domain feature focusing and a nonlinear sensor intensity change and dynamic trend adjustment mechanism, high-precision positioning of fire precursor and equipment fault is realized; a secondary screening mechanism and a dynamic threshold adjustment strategy are introduced, so that the accuracy of a fire signal and equipment anomaly detection result is remarkably improved; through construction of early detection reward loss, the penalty of early exposure of fire precursor signals or equipment faults is reduced, the early warning requirements of fire and abnormal equipment faults are met, and then the early warning accuracy and response efficiency of the building fire-fighting monitoring system are comprehensively improved.

The invention discloses a thyroid postoperative bleeding monitoring method and system, and relates to the technical field of medical monitoring. According to the method, the YOLO-Seg network model is adopted, and the SRCNN layer and the Blind Dconv layer are combined, so that the drainage tube area and the dressing area can be identified with high precision, the accuracy of area identification is improved, and the definition and the deblurring effect of the image are improved; the color, the category and the proportion information of the drainage liquid are accurately detected, so that powerful support is provided for judging whether the patient bleeding or not; by comprehensively analyzing the physiological index data, the overall health condition of the patient is more comprehensively evaluated, the health condition of the patient is accurately reflected through the physiological index score, the potential bleeding risk is found in time, a comprehensive, accurate and timely solution is provided for monitoring the thyroid postoperative patient, and safety and rehabilitation of the patient are effectively guaranteed.

The invention discloses a multi-task electromagnetic model based on a hybrid expert network, and belongs to a wireless communication technology. The model comprises a preprocessing module, a feature extraction module, a task output module and a pre-training-fine tuning learning strategy. The preprocessing module carries out standardization processing on the multi-source electromagnetic signals; the feature extraction module is based on a Transform structure, introduces a hybrid expert network to replace part of a traditional feedforward neural network, and dynamically selects an expert sub-network through a task specific routing mechanism; the task output module configures a special structure according to different task targets; in the pre-training stage, a mask auto-encoder is used for pre-training large-scale label-free data, and a downstream task is subjected to full-amount fine adjustment through small-scale label data. According to the method, multi-task collaborative learning and differential expression are realized, and the recognition performance, robustness and processing efficiency of the model in a complex electromagnetic environment are improved.

The invention discloses a drawing data extraction method based on multi-modal features, which comprises the following steps: S1, multi-modal data input and preprocessing: receiving multi-modal data, and preprocessing the multi-modal data; s2, multi-modal feature extraction: carrying out feature extraction on the data of each modal, and respectively generating a 256-dimensional feature vector of each modal; s3, feature fusion and alignment: performing multi-channel fusion and feature alignment on features generated by different modal feature generators; s4, training a drawing data detection model: carrying out model training by using the marked normal sample data and a multi-mode auto-encoder model; and S5, drawing data extraction: inputting new multi-modal data into a plurality of feature generators to generate multi-modal features, carrying out fusion to generate a C * 256 feature matrix, inputting the C * 256 feature matrix into the trained multi-modal auto-encoder, and carrying out data extraction.

The invention relates to a motor shell rusty spot recognition method, device and equipment and a computer readable medium. The method comprises the following steps: acquiring a first motor shell image before a salt spray test and a second motor shell image after the salt spray test; performing image registration on the first motor shell image and the second motor shell image to perform registration on a rusty spot area in the first motor shell image and a rusty spot area in the second motor shell image to obtain an effective motor shell image; performing difference detection on the effective motor shell images before and after the salt spray test, and extracting a newly added rusty spot area based on a difference detection result; constructing a rusty spot recognition model according to the newly added rusty spot area and the corresponding motor shell image; based on the rusty spot recognition model, the rusty spot type of the newly-added rusty spot area is recognized, and the area corresponding to the newly-added rusty spot area is calculated. According to the invention, the problem of low detection efficiency and accuracy caused by manual motor shell rusty spot recognition detection is solved.

The invention discloses a current transformer error prediction method and system, and relates to the technical field of online monitoring and error evaluation of current transformers, and the method comprises the steps: collecting signal data from equipment, and carrying out the preprocessing; carrying out dimension reduction processing on the preprocessed data features; and performing feature selection and prediction on the feature data after dimension reduction, and outputting a result. According to the invention, by combining the time domain, frequency domain and wavelet domain features of the current transformer monitoring signal, the multi-domain feature prediction model is constructed, and the capability of accurately identifying the operation condition of the current transformer is improved. And a CatBoost algorithm is used to evaluate the importance of the features, so that the feature interpretability of the model is improved, and the effectiveness of model training and the accuracy of prediction are ensured. And HistGB is selected as a regression model, so that the training speed is greatly improved, and the calculation cost and memory consumption are remarkably reduced when a large data volume is processed. According to the invention, quantitative analysis and timely prediction of the error of the high-voltage current transformer are realized.

The invention relates to a method for monitoring cerebral apoplexy, a device for monitoring cerebral apoplexy and a storage medium, and the method comprises the steps: collecting physiological parameters such as electroencephalogram (EEG), oxyhemoglobin saturation (SpO2), electrocardiogram (ECG) and blood pressure through wearable equipment or a family health monitoring system, and carrying out data preprocessing, feature extraction and anomaly detection; the EEG signals are processed by utilizing frequency domain characteristics and short-time energy analysis, and the cerebral apoplexy risk is identified by adopting a support vector machine (SVM) classification model. The system can analyze and detect abnormal conditions in real time, and sends out early warning information through a mobile phone App, a short message or a telephone to remind a user to see a doctor in time. The system comprises wearable equipment, a family health system, a hospital platform and a mobile health application, and is suitable for personal health monitoring and centralized monitoring of medical institutions. The real-time performance and accuracy of cerebral apoplexy monitoring are remarkably improved, the user experience is enhanced, the medical response speed is increased, and the method has wide application prospects and practical value.

The invention provides a driving behavior data analysis method based on end-cloud collaboration. The method comprises the following steps: 1) constructing a driving behavior image data set and a vehicle data set; 2) based on an improved Faster-RCNN model, carrying out image recognition on the driving behavior image data set, and obtaining behavior data; 3) transmitting the acquired vehicle data and the generated behavior data to a back-end server or a cloud end based on an API (Application Program Interface) provided by the platform; and 4) constructing a driving behavior scoring data set, performing driving safety scoring by using the decision tree model in combination with the data, generating a corresponding report according to a scoring strategy, and feeding back the report to a platform user through an event triggering mechanism.

The invention relates to an annotation processing method and device, a server and a storage medium. The annotation processing method comprises the following steps: performing information extraction on an image through at least two preset recognition models to obtain first description information of the image; performing first processing on the image to obtain a first processed image; and inputting the first description information of the image and the first processing image into a preset visual language model to obtain first annotation information of the image. According to the scheme provided by the invention, the accuracy and comprehensiveness of marking description can be improved, and the accurate recognition capability of the vehicle on the surrounding environment is improved.

The utility model discloses a flour conveying adjusting device which comprises a machine frame, a control unit, an auger conveying device and a material pipe, a sealing plate is arranged at the connecting position of the auger conveying device and the material pipe, and the sealing plate is connected with the end of the corresponding auger conveying device through an elastic reset assembly. A material level sensor is arranged in the end, corresponding to the material pipe, of the auger conveying device, and a displacement sensor is arranged on the outer side wall of the auger conveying device. And the accurate recognition capability of the blockage and the sensitivity of monitoring and early warning are improved.

The application belongs to the technical field of vehicle information anonymization detection, and particularly relates to a multi-type image anonymization annotation dataset construction and target coverage rate determination method. The method is based on a face and license plate image dataset with double annotation of theoretical anonymization region and anonymization features, introduces an anonymization feature extraction branch in an improved YOLOv5-L model and performs cross-modal feature fusion, combines a plurality of loss functions with theoretical anonymization region positioning loss as the core and a two-stage progressive training and difficult example mining mechanism, and realizes precise learning of the anonymization features. After normalizing the input anonymization image, the model inference obtains the theoretical anonymization region coordinates and maps them back to the original size, and through non-maximum suppression and matching of the IoU threshold, the region coverage rate is calculated to determine the positive detection, missed detection and statistical false detection rate. The application effectively overcomes the feature dependency failure and model robustness problem, and realizes high-precision, low-misjudgment anonymization detection and evaluation under various anonymization conditions.

The application discloses a method for identifying large-area fresh water and salt water lake types, which comprises the following steps: binarizing lakes according to the measured salt and fresh water types of the lakes; obtaining attribute data within a research area range; extracting time series values of attribute data of each lake basin within the research area range, and calculating the average value of each attribute data; counting the attribute differences between salt water and fresh water lakes, and selecting the attribute with significant difference as a key attribute; for each key attribute, searching for an optimal segmentation threshold of salt water and fresh water lakes within the value range of the key attribute; taking the optimal identification precision of each key attribute as a weight, and taking the salt and fresh water type identification result of each key attribute under the optimal segmentation threshold as a value to perform weighted average calculation, rounding off the weighted average calculation result, and identifying salt water and fresh water lakes according to the final calculation result. The application can realize high-precision identification of salt and fresh water lakes in a large area.

The invention discloses a subsynchronous harmonic compensation method and device of a power system. The method comprises the following steps: acquiring a real-time load current signal; based on a preset first window function, performing windowing processing on the real-time load current signal to obtain a first windowing current signal; carrying out peak detection on the first windowed current signal in a preset frequency range to obtain respective frequencies, amplitudes and phases of a plurality of subsynchronous harmonics; judging whether to maintain a first window function based on the frequencies and amplitudes of the plurality of subsynchronous harmonics; based on the respective frequencies, amplitudes and phases of the plurality of subsynchronous harmonics, sine wave signals corresponding to the plurality of subsynchronous harmonics are generated; and compensating the real-time load current signal through a unified power quality regulator based on the sine wave signals corresponding to the plurality of subsynchronous harmonics. According to the invention, the technical problems of low subsynchronous harmonic detection precision and weak separation capability of the existing unified power quality conditioner are solved.

An object recognition method according to one embodiment of the present disclosure includes: a step of obtaining image data of an image including feature information indicating features of an object; and a step of recognizing the object included in the image based on the feature information. The image data is obtained by capturing the image with a first camera, the first camera having an image sensor and a filter array, the filter array being arranged in an optical path of light incident on the image sensor, and including a plurality of light-transmitting filters arranged two-dimensionally along a surface intersecting the optical path, the plurality of filters including two or more filters having mutually different wavelength dependencies of light transmittance, and the light transmittance of each of the two or more filters having a local maximum value in a plurality of wavelength regions.

The utility model provides a kind of jacking rotary code scanning device, it is related to battery production technical field.Jacking rotary code scanning device includes first driving part, mounting piece, transmission part and second driving part.The output end of first driving part is connected with mounting piece, for driving mounting piece to lift, the number of transmission part is multiple, multiple transmission parts are movably arranged in mounting piece and are sequentially transmission connection, the output end of second driving part is connected with any one transmission part, for driving multiple transmission parts to rotate, transmission part is used to drive battery to rotate, to this is conducive to identification mechanism accurate identification battery, and by driving battery to rotate can effectively avoid the problem that identification mechanism cannot be identified due to battery surface reflection light influence identification mechanism cannot be identified.Obviously, the jacking rotary code scanning device provided by the utility model embodiment is high in degree of automation, and significantly improves identification success rate and production efficiency.

This invention relates to the field of spectral analysis and identification of genuine and counterfeit Chinese medicinal materials, and discloses a method for identifying counterfeit Cordyceps sinensis based on multispectral fusion. This method first collects multispectral image data of samples under standard conditions; eliminates equipment and environmental noise through standardized preprocessing to establish a reliable data foundation; then, it fuses extracted spectral derivative features and spatial texture features to construct a lightweight neural network identification model. Its core innovation lies in introducing a physical interference stripping model, which can actively correct spectral distortion caused by moisture and mold, obtaining intrinsic spectra to improve model robustness; and through dynamic verification and incremental learning mechanisms, the model can continuously evolve, effectively identifying new counterfeiting methods. Finally, the entire model is integrated into a multispectral portable terminal, achieving rapid screening within seconds and outputting a complete report including authenticity determination, type inference, and confidence level. Ultimately, this solves the industry problems of low efficiency, reliance on experience, and inability to be applied on-site by traditional methods.

This image processing device is provided with: an image acquisition unit that acquires a target image in which a plurality of fibers are captured; a first region division unit that generates a region division result for each individual, in which each fiber included in the target image is detected, using the region division model for each individual that has been learned; a second region division unit that generates a region division result for each type in which a region in which a plurality of fibers are captured is recognized from the target image, using the region division model for each type that has been learned; a region correction unit that corrects the result of the region division for each individual by means of the result of the region division for each type; and a result output unit that outputs a correction result of the region division result for each individual.

The present invention discloses a servo system and its initial angle identification method, device, motor controller and medium, wherein the method comprises: determining a first angle of a motor of the servo system; obtaining a first current given instruction, and controlling the motor to rotate in a first direction according to a preset angular frequency according to the first current given instruction; in the process of the motor rotating in the first direction, determining a first angle difference between the actual electrical angle of the motor and the electrical angle corresponding to the first current given instruction; determining the initial angle of the motor according to the first angle and the first angle difference. Thus, the initial angle of the motor in the servo system can be identified with high precision.