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

31results about How to "Integrated precision" patented technology

Vehicle collision risk early warning method, system, equipment and medium

The invention provides a vehicle collision risk early warning method, system and device and a medium, and belongs to the technical field of vehicle safety control. The method comprises the steps that surrounding environment information of a vehicle is collected in real time through a vehicle built-in sensor and a GPS module; a multi-sensor fusion device technology is used to carry out comprehensive processing on surrounding environment information of a vehicle, and an environment perception model is generated after data integration; based on the environment perception model, target identification is carried out by using a computer vision technology based on a convolutional neural network so as to identify travelers, vehicles, traffic signs and obstacles; performing behavior prediction based on a deep learning model to predict a behavior track of a front vehicle or pedestrian; evaluating a potential collision risk in the current scene in combination with the target recognition result and the behavior prediction result; when a potential collision risk is detected, an early warning signal is sent to a driver, and when the driver does not take corresponding measures, active safety measures are automatically triggered to avoid collision or reduce accident damage.
Owner:浪潮智慧科技有限公司 +1

Typhoon wind speed prediction method based on Mamba-ASPP and R-TCN

The invention discloses a typhoon wind speed prediction method based on Mamb-ASPP and R-TCN. The method comprises the following steps: obtaining a satellite cloud picture data set and carrying out preprocessing; a Mama-ASPP module is constructed for spatial feature extraction, an R-TCN module is constructed for time feature extraction, a global spatial feature map and a time feature map of an input image are obtained and spliced, the spliced image is put into an ITSAM module for comprehensive feature processing, and a predicted value of the typhoon wind speed is obtained; integrating the three modules into an initial wind speed prediction model, and optimizing and training hyper-parameters of the initial wind speed prediction model by using a seagull algorithm optimized by an adaptive genetic algorithm to obtain a target wind speed prediction model; and inputting the image data of the previous hour into the target wind speed prediction model, and predicting the typhoon wind speed of the next hour. According to the method, by integrating different time and space scale feature information, the complexity of typhoon wind speed change is dealt with, and the typhoon prediction efficiency and accuracy are effectively improved.
Owner:NANJING UNIV OF INFORMATION SCI & TECH

Tooth three-dimensional digital model registration and tooth neck line extraction method based on oral scanning and CBCT

The invention discloses a tooth three-dimensional digital model registration and tooth neck line extraction method based on oral scanning and CBCT, and relates to the field of tooth three-dimensional digital model registration. Comprising the following steps: respectively acquiring three-dimensional oral images of oral scanning and CBCT (cone beam computed tomography), and automatically segmenting three-dimensional structures of dental crowns and teeth by utilizing pre-trained V-Net and ToothNet image segmentation models. Alignment and registration of two modal images are realized through a feature extraction network, an information exchange network and a rigid transformation network. The feature extraction network extracts local and global features by using far-point sampling and a dynamic graph convolutional neural network, the self-attention and cross-attention module performs feature enhancement and information exchange between the same mode and different modes, and based on a registration result, a tooth neck line is further extracted and optimized. According to the method, the tooth three-dimensional digital model can be quickly and accurately registered, a high-precision tooth neck line extraction result is provided, the manpower and time cost of a medical institution can be reduced, and the medical efficiency is improved.
Owner:CHONGQING UNIV OF POSTS & TELECOMM +1

Certificate management method and system based on pre-trained large language model, and readable medium

The invention belongs to the technical field of certificate management of intelligent application, and relates to a certificate management method and system based on a pre-trained large language model and a readable medium, and the method comprises the following steps: collecting certificate information of different subsidiary companies, different departments and different formats, and carrying out the standardization processing of the certificate information; storing the standardized certificate information in a distributed storage system to support storage and quick retrieval of a large amount of certificate information; carrying out feature extraction on the certificate information of the distributed storage system, and establishing association between the certificate information; recommending related certificate information based on the association between the certificate information according to the demands and historical behaviors of the user; and checking whether certificate information in the distributed storage system accords with company policies and industry standards or not, performing risk assessment, and providing a corresponding solution. The method aims at providing efficient and personalized certificate management services for large group companies and second-level and third-level companies thereof by introducing a pre-training large language model technology.
Owner:CRSC INFORMATION IND CO LTD

Grid fault rapid diagnosis method based on artificial intelligence technology

The invention discloses a power grid fault rapid diagnosis method based on an artificial intelligence technology, and relates to the field of power grid fault diagnosis, and the method comprises the steps: S1, carrying out the feature extraction of a power grid fault based on quantum calculation; s2, constructing a power grid knowledge graph; s3, mapping entity nodes in the power grid knowledge graph to a quantum state according to key attribute information of the entity nodes, and designing quantum gate operation for encoding according to a relationship among the entity nodes in the power grid knowledge graph; s4, performing fault diagnosis; and S5, performing feedback and optimization based on the power grid knowledge graph. According to the rapid power grid fault diagnosis method based on the artificial intelligence technology, the high efficiency of the quantum calculation in the aspects of data processing and feature extraction and the advantages of the knowledge graph in the aspects of knowledge representation and reasoning are brought into full play through the power grid fault diagnosis method based on fusion of the quantum calculation and the knowledge graph; rapid and accurate diagnosis of power grid faults can be realized, and a guarantee is provided for safe and stable operation of a power grid.
Owner:INNER MONGOLIA FINANCE AND ECONOMICS UNIVERSITY

Basin unit debris flow susceptibility evaluation method based on multi-model Stacking integration

The invention discloses a drainage basin unit debris flow susceptibility evaluation method based on multi-model Stacking integration. The method comprises the following steps: constructing a debris flow disaster data set; constructing a Stacking ensemble learning model which comprises four base models, a feature stacking layer, an attention layer, a mixed feature layer, a feature splicing layer and a meta-model; a Stacking integrated learning model is trained; predicting the debris flow disaster probability of the drainage basin unit to be identified; and performing debris flow susceptibility evaluation. According to the method, multi-source data can be efficiently fused, and the mixed probability vector and the original sample are creatively spliced, so that the meta-model can simultaneously utilize the intrinsic characteristics of the original data and the prediction information of the base model in the training process; interference of irrelevant information and dependence on the quality of a base model can be reduced, key information in data can be captured more comprehensively, and the explanation capacity and prediction precision of the model on complex data are enhanced.
Owner:CHENGDU UNIVERSITY OF TECHNOLOGY +1

Book resource sharing and management platform based on cloud computing

The invention discloses a book resource sharing and management platform based on cloud computing, relates to the technical field of book resource management, and realizes scientific quantification and durability dynamic analysis of book loss by combining a plurality of algorithms in a data processing module and a cloud computing analysis module. The data processing module constructs an environmental influence factor EHF and an annual aging factor AF through data cleaning and normalization processing, and effectively evaluates the long-term influence of humidity and time on books. Meanwhile, texture and color feature extraction is performed on the image data based on a visual algorithm, and a cover loss score CS and an internal paper wear score IS are generated. And the cloud computing and analyzing module dynamically computes a single borrowing loss rate Ds, an accumulated loss value Dt and a durability score Sr by using the comprehensive loss data set. The module generates a maintenance prompt through the evaluation of the dynamic loss rate and the long-term loss trend, and preliminarily evaluates the use condition of the book.
Owner:HENAN AGRICULTURAL UNIVERSITY

Human-computer interaction processing method and device, equipment and storage medium

The invention provides a man-machine interaction processing method and device, equipment and a storage medium, and the method comprises the steps: obtaining an image of a user interface according to a preset frequency; obtaining an operation instruction of a user; inputting the image into an image processing model obtained by pre-training, performing content extraction on the image, and generating structured information corresponding to the image; inputting the image, the structured information and the operation instruction into a multi-modal large model obtained by pre-training, and generating a task statement corresponding to the operation instruction; and executing the task statement for the user interface, and outputting an execution result to the user. According to the man-machine interaction processing method, the multi-modal information can be accurately integrated, reasoned and decided, man-machine interaction under multiple modes is achieved, meanwhile, the man-machine interaction processing method provided by the embodiment of the invention can be executed for multiple times so as to carry out man-machine interaction processing on complex operation requirements of a user, and the user experience is improved. The man-machine interaction processing method provided by the embodiment of the invention can also be suitable for complex industrial scenes.
Owner:WUXI XUELANG DIGITAL TECH CO LTD

Intention recognition data query method and system based on multilayer model

The invention discloses an intention recognition data query method and system based on a multilayer model, and the method comprises the steps: mapping a data source needing to be queried into a basic model; on the basis of the basic model, creating a logic model based on a user-defined SQL query statement; a user selects and combines one or more basic models and logic models to create an intention model; receiving model query parameters including query fields, whether aggregation exists or not, filtering conditions and the like, and distinguishing query targets through model types; for query of the basic model and the logic model, automatically generating SQL query statements according to model query parameters; and for query of the intention model, dynamically routing to different models for query according to model query parameters and an addressing rule algorithm, and if multiple models are addressed, performing associated query according to dimensions. According to the method and the system, efficient mapping and execution from a data source to a query intention are realized by constructing a three-layer architecture of a basic model, a logic model and an intention model.
Owner:CHINA UNITECHS

Multi-dimensional data fusion and intelligent analysis system for smart park

The invention relates to the technical field of data analysis, in particular to a multi-dimensional data fusion and intelligent analysis system for a smart park, which is characterized in that after various types of data are artificially divided into time sequence features, spatial features and attribute features, a controller respectively performs similarity calculation on the three features and then performs fusion; the time sequence features and the spatial features can cover most of data, the attribute features are used for capturing data which are not suitable for time and space distribution, and regular paths are introduced into the time sequence, so that unified path planning can be utilized to avoid data reference errors caused by path asynchronization and further avoid fusion failures; fusion is carried out by fully considering multiple features such as time, space and attributes of the data, information loss or wrong fusion caused by data isomerism is avoided, so that the accuracy of data fusion is improved, the capture capability of the data fusion method on complex data relationships is higher, and the data fusion efficiency is improved. And the fused data can more truly reflect the actual operation state of the park.
Owner:FUJIAN POLYTECHNIC OF INFORMATION TECH +1

Forgery audio detection method, device, equipment and medium

The invention relates to the technical field of artificial intelligence and financial science and technology, and provides a forged audio detection method, device and equipment and a medium, and the method comprises the steps: carrying out the data preprocessing and feature extraction of to-be-detected audio data, so as to obtain an original feature embedding vector; performing feature compression on the original feature embedded vector to obtain a compressed embedded vector; key feature extraction is carried out on the original feature embedding vector through an attention statistics pool layer and a multi-layer perceptron to obtain a key feature embedding vector; fusing the compressed embedded vector and the key feature embedded vector through a multi-head attention mechanism to obtain a fused feature vector; and detecting the fused feature vector through the classification head to determine whether the to-be-detected audio data is a forged audio or not. Through the technical scheme of the invention, the recognition precision of the forged audio is improved, and meanwhile, better generalization ability is shown on data in different fields.
Owner:PING AN TECH (SHENZHEN) CO LTD

Method and device for processing data by vehicle end rule engine and electronic equipment

The invention provides a method and a device for processing data by a vehicle end rule engine and electronic equipment, in the method, a user can configure a data processing rule through a graphical interface or a programming interface, the rule configuration is more flexible, and the user can interact with an external system so as to obtain external data of the external system; the method comprises the following steps: performing parallel aggregation calculation processing on external data and sensor data in a preset time window by adopting a multi-thread calculation technology, and meanwhile, establishing a data relation graph of the data after aggregation calculation, the external data, the sensor data and static data; according to the method, the optimal processing path of the processing path corresponding to the data processing rule in the data relation graph is determined by adopting the shortest path algorithm, and finally, data processing is performed according to the optimal processing path, so that the data processing efficiency can be greatly improved, the efficient processing of the real-time data stream is realized, and the data processing accuracy is improved.
Owner:NEUSOFT REACH AUTOMOBILE TECH (SHENYANG) CO LTD

A method and system for automated generation of TSN network configuration based on reflective agents

ActiveCN121441738BImplement dynamic updatesAchieve continuous evolutionBiological modelsTransmissionResource informationDistributed computing
This invention relates to a method and system for automated generation of TSN network configurations based on a reflective agent. The system comprises a network state perception and intent reflection module, a YANG model configuration parameter and structure decision module, and a YANG model intelligent generation module. The network state perception and intent reflection module is responsible for structurally perceiving network state and resource information. The YANG model configuration parameter and structure decision module integrates multi-source knowledge and contextual information, intelligently generating deployable TSN scheduling parameters and YANG configuration structures based on business needs. The YANG model intelligent generation module automatically generates the corresponding YANG configuration model based on the optimized results, performs syntax and semantic verification, and then distributes it to the target device, achieving automated deployment of network configurations. This invention significantly improves the intelligence level and deployment automation capability of TSN network configuration by introducing reflective reasoning and feedback optimization techniques.
Owner:STATE GRID HUBEI ELECTRIC POWER RES INST

A forestation configuration method and system based on geographic data

The present invention relates to the field of afforestation configuration technology, and is a afforestation configuration method and system based on geographic data, and the specific method includes: collecting initial geographic data of historical afforestation areas; performing feature selection processing and data dimension reduction processing on the initial geographic data; classifying afforestation design patterns of different afforestation areas to be selected, and evaluating the applicability of afforestation design patterns; updating the alternative afforestation design patterns of afforestation areas under different geographical conditions to the afforestation personnel design platform, and automatically providing suggestions to afforestation personnel. The present invention solves the problem in the prior art that the requirements of afforestation engineering management involve many aspects such as plot segmentation, garden landscaping, basal fertilizer application, shading, support, horizontal belt landscaping, working roads, etc., which leads to a huge workload when the design content is complicated.
Owner:ZHEJIANG FOREST RESOURCES MONITORING CENT (ZHEJIANG FORESTRY SURVEY PLANNING & DESIGN INST) +1

Method and device for constructing electronic commerce large model, and medium

PendingCN119443269ARaise the level of digitizationreduce riskForecastingInference methodsStreaming dataFinancial transaction
The invention discloses an electronic commerce large model construction method and device and a medium, and relates to the technical field of electronic commerce. The method comprises the steps of obtaining historical transaction data, logistics data and unstructured data, and extracting features of the unstructured data through natural language processing; according to the historical transaction data and the logistics data, a sales prediction model and a logistics timeliness prediction model are constructed through Hadoop, and an e-commerce model is established based on an e-commerce industry database and an AI large model; performing cross validation and parameter optimization on the sales prediction model and the logistics timeliness prediction model, and integrating with an electronic commerce model to obtain an electronic commerce large model; and performing micro-servitization on the electronic commerce large model, and configuring a service discovery mechanism to enable micro-services to communicate with each other. According to the method, a comprehensive electronic commerce large model is formed, and comprehensive data analysis and decision support can be provided for enterprises.
Owner:INSPUR ZHUOSHU BIG DATA IND DEV CO LTD

Visible light-infrared target detection method based on structured feature extraction and dynamic weight prediction

The invention discloses a visible light-infrared target detection method based on structured feature extraction and dynamic weight prediction. The method comprises the following steps: carrying out target detection through a visible light-infrared target detection model based on structured feature extraction and dynamic weight prediction; the target detection model comprises a thermal radiation-motion feature extraction module, a multi-granularity visual representation module and a collaborative weight decision module; the thermal radiation-motion feature extraction module is used for capturing thermal radiation information in an infrared image and motion features of a target so as to generate infrared feature representation; the multi-granularity visual representation module is used for performing structured multi-granularity visual feature extraction on the visible light image so as to construct visible light feature representation; and the collaborative weight decision module is used for performing cross integration of infrared feature representation and visible light feature representation through a dynamic weight prediction mechanism. The target detection model provided by the invention has better performance in a visible light-infrared target detection task.
Owner:YELLOW RIVER CONSERVANCY TECHN INST +1

Afforestation configuration method and system based on geographic data

The invention relates to the technical field of afforestation configuration, in particular to an afforestation configuration method and system based on geographic data, and the method specifically comprises the steps: collecting initial geographic data of a historical afforestation region; performing feature selection processing and data dimension reduction processing on the initial geographic data; classifying the afforestation design modes of different afforestation areas to be selected, and performing afforestation design mode adaptability evaluation; and the alternative afforestation design modes of the to-be-afforestation area under different geographical conditions are updated to an afforestation personnel design platform, and suggestions are automatically provided for afforestation personnel at the same time. According to the method, the problem that in the prior art, due to the fact that the requirements of afforestation engineering management relate to land parcel segmentation, garden landscaping, base fertilizer application, shading, supporting, horizontal belt soil preparation, operation ways and other aspects, the workload is huge when the design content is complex is solved.
Owner:ZHEJIANG FOREST RESOURCES MONITORING CENT (ZHEJIANG FORESTRY SURVEY PLANNING & DESIGN INST) +1

Multi-source marine geological information fusion and 3D visualization modeling method

The present invention relates to the technical field of information processing and three-dimensional modeling, and specifically to a method for multi-source marine geological information fusion and three-dimensional visualization modeling, comprising the following steps: S1: collecting marine geological data from multiple sources; S2: preprocessing the collected data; S3: extracting marine geological features from the preprocessed data; S4: multi-dimensionally fusing the features extracted in S3; S5: constructing a marine geological three-dimensional model based on the fused data in S4 using a three-dimensional grid construction algorithm; S6: optimizing the constructed three-dimensional model; S7: visualizing the optimized three-dimensional model to generate an interactive three-dimensional display interface. The present invention improves the accuracy, information integrity and user interactive analysis capabilities of the marine geological model through multi-source data fusion, refined three-dimensional model construction and efficient visualization processing, and provides reliable technical support for geological research and engineering applications.
Owner:QINGDAO INST OF MARINE GEOLOGY

Gene library acquisition method for improving coverage degree of single cell amplification genome

The invention relates to the field of biological information, and discloses a gene library acquisition method for improving single cell amplification genome coverage. Aiming at the problem of insufficient single cell amplification genome coverage, the gene library acquisition method provides a strategy of composite library construction and optimized data integration, and comprises the following steps: constructing at least one original text library and one sub-library with different average fragment lengths for the same single cell sample, and sequencing; performing Barcode resolution and effective single cell identification on the sequencing data through a customizable sequence feature detection method; and further, according to the identification result of the effective single cell, identifying, collecting and merging all sequencing sequence data of the same single cell, generating an integrated sequence data set of the corresponding single cell, and obtaining gene library data of the single cell. The SAG coverage degree of the obtained gene library data is high, and the method has high reliability and application value when being used for genome analysis of the single cell downstream.
Owner:MOBIDROP (ZHEJIANG) CO LTD +1

Data processing method and system applied to three-dimensional medical teaching system

PendingCN119107413AIncreased operational complexitySimplify operational complexityData processing applications3D modellingMedical knowledgeHuman–computer interaction
According to the data processing method and system applied to the three-dimensional medical teaching system provided by the invention, based on teaching composition interaction of the basic teaching composition application and the derivative teaching composition application, basic medical knowledge and medical contents in advanced or specific fields can be visually and vividly displayed; and the learning experience and the understanding depth of students are effectively enhanced. The first teaching composition interaction data and the second teaching composition interaction data are loaded and integrated in real time, dynamic generation of platform interaction resources is promoted, a teaching interface can reflect latest teaching content and interaction results in real time, and therefore the timeliness and interactivity of teaching are remarkably improved. Therefore, by optimizing the data processing flow and the resource organization mode, the operation complexity of the three-dimensional medical teaching system is simplified, and the teaching quality and efficiency are greatly improved.
Owner:SHANGHAI BOHAO INFORMATION TECH CO LTD

Manuscript generation method and device, equipment and storage medium

The invention relates to the technical field of text processing, and discloses a manuscript generation method and device, equipment and a storage medium. The method comprises the following steps: constructing a template library containing a plurality of template frames and a knowledge base containing a plurality of domain knowledge bases; obtaining use features of a user and each template frame, receiving a free prompt input by the user, and constructing a task feature vector; matching a target template frame based on the first sorting model, and generating a task instruction based on a standard prompt word of the target template frame; if the matching target template frame does not exist, generating a task instruction based on the free prompt word; matching a target knowledge fragment from the knowledge base based on a second sorting model; and forming a large model generation manuscript called by a generation instruction in combination with the target knowledge fragment and the task instruction, correcting the manuscript in response to interaction between a user and the manuscript until a finalized manuscript is generated, and collecting interaction data to perform quality evaluation on the finalized manuscript. According to the method, high-quality, personalized and automatic generation of the manuscripts in the specific field is realized.
Owner:CSC FINANCIAL CO LTD

A method for assessing debris flow susceptibility in watershed units based on multi-model stacking integration

ActiveCN119312204BGuaranteed accuracyIntegrated precisionData processing applications
The present invention discloses a method for assessing the debris flow susceptibility of a watershed unit based on multi-model Stacking integration, comprising the steps of: constructing a debris flow disaster data set; building a Stacking integrated learning model, including four base models, a feature stacking layer, an attention layer, a mixed feature layer, a feature splicing layer and a meta-model; training the Stacking integrated learning model; predicting the probability of debris flow disasters in the watershed unit to be identified; and assessing the debris flow susceptibility. The present invention can efficiently fuse multi-source data, and creatively splice the mixed probability vector with the original sample, so that the meta-model can simultaneously utilize the intrinsic features of the original data and the prediction information of the base model during the training process, can reduce the interference of irrelevant information and the dependence on the quality of the base model, can more comprehensively capture the key information in the data, and enhance the model's ability to interpret complex data and the prediction accuracy.
Owner:CHENGDU UNIVERSITY OF TECHNOLOGY +1

Power operation and maintenance information mining method and system based on inspection business

The embodiment of the invention relates to the technical field of electric power data analysis, in particular to an inspection service-based electric power operation and maintenance information mining method and system, which can comprehensively and accurately integrate operation and maintenance information by distinguishing structured and unstructured parts in target electric power operation and maintenance information. Operation and maintenance inspection semantic vectors and equipment operation index vectors are mined for different types of information, and respective information values are fully mined. After vector interaction, operation and maintenance inspection decision vectors are obtained, then fault weights are obtained through discriminant analysis, and deep fusion and analysis of multi-source information are achieved. According to the mode, the abnormal operation and maintenance information is accurately judged, the operation and maintenance efficiency and precision are improved, the manual troubleshooting workload is reduced, the fault risk of a power system is effectively reduced, the intelligence and reliability of power operation and maintenance are enhanced, and the stability of power supply is guaranteed.
Owner:JIANGSU ELECTRIC POWER INFORMATION TECH

A typhoon wind speed prediction method based on Mamba-ASPP and R-TCN

ActiveCN119274083BEfficient fusionBalanced weightImage enhancementImage analysis
The present invention discloses a typhoon wind speed prediction method based on Mamba-ASPP and R-TCN, and the steps are as follows: obtaining a satellite cloud image data set and performing preprocessing; constructing a Mamba-ASPP module for spatial feature extraction, constructing an R-TCN module for temporal feature extraction, obtaining a global spatial feature map of an input image and a temporal feature map splicing, placing an ITSAM module for comprehensive feature processing, and obtaining a predicted value of the typhoon wind speed; integrating the three modules into an initial wind speed prediction model, optimizing and training the initial wind speed prediction model hyperparameters with a seagull algorithm optimized by an adaptive genetic algorithm, and obtaining a target wind speed prediction model; inputting the image data of the previous hour into the target wind speed prediction model, and predicting the typhoon wind speed of the next hour. The method of the present invention effectively improves the efficiency and accuracy of typhoon prediction by integrating feature information of different time and space scales to cope with the complexity of typhoon wind speed changes.
Owner:NANJING UNIV OF INFORMATION SCI & TECH

A kind of electric power synchronous meter reading method and meter reading device

The invention discloses a method and a device for synchronous electric power meter reading, which relates to the technical field of electric power meter reading, and solves the technical problems of imperfect time calibration and updating mechanism of meter reading equipment and lack of systematic multi-dimensional verification and analysis process for periodically collected data. The invention uses a built-in high-precision clock module, calibrates according to quartz crystal pulses every second and continuously interacts with an authoritative time service source, thereby improving the operation and maintenance repair speed and the accuracy of the time dimension of electric power data, and finely controls multiple links of startup and network access to ensure stable network connection, avoids data reporting loss like automatic reconnection of remote office VPN, supports continuous and stable transmission of electric power data to a main station, verifies and classifies collected data in multiple dimensions, provides a solid and reliable data foundation for electric power allocation and cost accounting, and intelligently monitors system upgrades and parameter changes, and performs initialization on demand priority. The whole process improves equipment stability and maintainability, ensures stable operation of meter reading equipment, and reduces failure downtime.
Owner:CHANGSHA SHEWEI METER INFORMATION TECHNOLOGY CO LTD

Camera shell detection method and detection system based on machine vision

The invention relates to the technical field of intelligent detection, and discloses a camera shell detection method and system based on machine vision, and the method comprises the steps: setting a plurality of image collection devices, collecting the image data of a to-be-detected shell, and obtaining a feature vector; splitting the image data chain into a plurality of sub-data chains; performing defect identification on the image data in each sub-data chain, and determining a defect area and a defect depth through contour overlapping degree de-duplication; obtaining surface detection evaluation of the to-be-detected shell according to the defect area and the defect depth; performing constant-temperature heating on the to-be-detected shell, and acquiring an infrared image of the to-be-detected shell to determine internal detection evaluation of the to-be-detected shell; and obtaining an overall defect value of the to-be-detected shell according to the surface detection evaluation and the internal detection evaluation, and performing detection early warning. By fusing the surface and internal detection evaluation values, quality evaluation and early warning are realized, and the problems of high omission ratio and low manual detection efficiency of traditional detection are solved.
Owner:SHENZHEN XIN MAO XIN IND CO LTD

Multi-screen video synchronization method and synchronization system

The present invention provides a multi-screen video synchronization method and synchronization system, which relates to the field of video processing technology, including obtaining time series by decoding multiple video streams, intelligently detecting and predicting time deviations, and accurately correcting deviations using a pre-trained neural network model to ensure the synchronization accuracy of each video stream. The method includes: obtaining the decoded time series of video frames, detecting and predicting time points that do not meet preset deviations, correcting deviations based on predicted values, integrating the corrected time series, and realizing accurate splicing and synchronous playback of video frames. The present invention realizes seamless and accurate synchronous playback of multi-screen video content through efficient time deviation detection, intelligent deviation prediction and correction, and accurate time series integration and video frame splicing.
Owner:SICHUAN GUANGXIN TIANXIA MEDIA CO LTD

Multimodal image semantic segmentation method and system based on cross-level guided fusion

The present invention discloses a multimodal image semantic segmentation method and system based on cross-level guided fusion, which utilizes multimodal data of RGB image information and additional modal image information such as depth information, thermal imaging information and spectral information, fuses multimodal information and uses it for image semantic segmentation tasks. The method first converts the additional modal image into the same size and visual range as the RGB image, and then uses a dual-branch encoder based on the Transformer basic block to extract RGB modal and additional modal features, and uses a cross-level guided fusion module to guide the multimodal fusion of high-level features in shallow, middle and high-level features using low-level features, and finally uses a decoder to decode the multi-level fusion features, and outputs the image category pixel by pixel, which solves the problem of poor semantic segmentation performance in complex scenes; the technical solution of the present invention can be applied to complex scenes such as robot intelligent sorting and drone intelligent inspection.
Owner:HUNAN UNIV

Multi-modal sensing information fused embedded large model decision control method and system

PendingCN121956559AAchieve embedded lightweight deploymentIntegrated precisionAdaptive controlDecision controlDecision strategy
The invention provides an embedded large model decision control method and system fusing multi-modal sensing information, and relates to the field of intelligent decision control, and the method comprises the steps: receiving multi-modal data of a space non-cooperative target, carrying out the preprocessing and feature extraction of the multi-modal data, and obtaining a multi-modal sensing information fusion model; deep semantic fusion is realized through unified semantic space projection and a dynamic confidence weighting strategy, and a unified environment representation vector is output; constructing an input representation fusing a time sequence context, domain knowledge Prompt and a sample example, performing target intention recognition and risk level evaluation in stages through a lightweight deployed large model, and generating a structured decision strategy; a structured decision strategy is converted into a spacecraft motion instruction, and a target control sequence is solved through a designed simplified MPC controller. According to the method, the non-cooperative target sensing precision and the data utilization efficiency can be improved, the autonomous intelligent decision-making capability of the terminal is enhanced, and safe and efficient capture and on-orbit service tasks are supported.
Owner:BEIHANG UNIV

Intelligent packaging production line automation control method and system

The application relates to the technical field of intelligent packaging production line automatic control and multi-station collaborative regulation, and discloses an intelligent packaging production line automatic control method and system, the method comprises the following steps: collecting visual and force feedback data, fusing and preprocessing to obtain multi-station integration information, identifying an abnormal mode, determining product switching, and generating a parameter adjustment scheme; issuing a regulation instruction to correct an execution parameter and analyzing a collaborative deviation; collecting updated data to analyze a deviation value and iteratively optimizing the parameter; executing action coordination and monitoring packaging completeness to obtain an adaptive stable state of the production line. The method realizes multi-station working condition collaborative sensing, intelligent adaptation of product switching and dynamic correction of parameters, solves the problems of response lagging, parameter regulation rigidity and insufficient packaging quality stability of a traditional packaging production line, and improves the operation efficiency and finished product qualification rate of the intelligent packaging production line.
Owner:ZHEJIANG MINGLI MASCH TECH CO LTD