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

570results about How to "Avoid subjectivity" patented technology

Visual project management method and system based on task tracking

The invention provides a visual project management method and system based on task tracking, and relates to the technical field of project management, and the method comprises the steps: collecting project risk information through employing a context awareness technology, generating a risk propagation network, and calculating a risk diffusion coefficient; performing hierarchical analysis on project requirements based on the risk diffusion coefficient to obtain a task decomposition scheme, constructing a task transmission chain based on the task decomposition scheme and generating a visual task relation graph, and generating a task state transition rule and an alternative task path based on an early warning state of a task node, meanwhile, constructing a task elastic scheduling mechanism and storing the task elastic scheduling mechanism into a rule database; and identifying a preorder task node and a post-order task node on the task transmission chain, sending the task reconstruction instruction to the preorder task node and the post-order task node to realize dynamic task adjustment, updating the risk propagation network according to an adjustment result, and optimizing the risk diffusion coefficient.
Owner:STATE GRID SIJI DIGITAL TECH (BEIJING) CO LTD +1

Vehicle illegal parking detection system based on patrol shooting of unmanned aerial vehicle

The invention discloses a vehicle illegal parking detection system based on unmanned aerial vehicle inspection shooting, and relates to the technical field of vehicle illegal parking inspection. The vehicle illegal parking detection platform is in communication connection with an unmanned aerial vehicle image acquisition module, an image preprocessing module, a vehicle detection analysis module, an illegal parking judgment module, a license plate recognition module and an early warning notification module, and all the modules are in electric signal connection. According to the method, a large area can be rapidly covered through unmanned aerial vehicle inspection shooting, real-time monitoring and detection of vehicle illegal parking behaviors are achieved, compared with a traditional manual inspection mode, unmanned aerial vehicle inspection greatly shortens the detection period, the detection accuracy is remarkably improved, and the detection efficiency is improved through a high-precision image recognition algorithm. The vehicle illegal parking detection system can accurately recognize the license plate number, the vehicle type and the illegal parking position, and subjectivity and errors of manual judgment are effectively avoided.
Owner:NANJING TUOHENG UNMANNED SYST RES INST CO LTD

Traditional Chinese medicine abnormal state recognition method, system and equipment and storage medium

The invention provides a traditional Chinese medicine abnormal state recognition method, system and device and a storage medium, and relates to the field of abnormal recognition. The method comprises the following steps: acquiring product information of a to-be-detected traditional Chinese medicine, and determining a data acquisition standard of the to-be-detected traditional Chinese medicine based on the product information of the to-be-detected traditional Chinese medicine and a preset detection rule; based on a data acquisition standard of the traditional Chinese medicine to be detected, inputting the image data into a preset image anomaly detection model to obtain a preliminary image anomaly region; performing feature fusion on the visual feature data and the spectral feature data to obtain an abnormal feature vector; and inputting the abnormal feature vector into a preset abnormal state judgment model, outputting abnormal state data, and performing abnormal state judgment based on the abnormal state data and a preset abnormal state judgment threshold to obtain an abnormal state detection result of the to-be-detected traditional Chinese medicine. Through the scheme, the abnormal state of the traditional Chinese medicine can be identified timely and accurately.
Owner:CONITECH (BEIJING) MANAGEMENT SOFTWARE CO LTD

Tunnel surrounding rock multi-mode intelligent fine grading method and system

The invention belongs to the field of tunnel engineering, and particularly discloses a tunnel surrounding rock multi-mode intelligent fine grading method and system, and the method comprises the steps: 1, constructing a sample library which comprises parameters while drilling of a tunnel face, a digital image and a corresponding tunnel face surrounding rock grade; 2, constructing a while-drilling parameter refined grading characteristic system; 3, constructing a surrounding rock intelligent grading model based on multi-modal data fusion; 4, dividing the rock mass in front of the tunnel face into a plurality of three-dimensional space unit bodies according to the tunnel contour and blast hole distribution characteristics; and 5, based on the surrounding rock intelligent grading model, finely and intelligently judging the grade of the tunnel face surrounding rock. According to the method, the intelligent surrounding rock grading model based on the multi-modal data fusion theory is constructed, the refined intelligent judgment of the tunnel face surrounding rock grade is realized, the precision and robustness of the refined intelligent judgment of the surrounding rock grade of the model are improved, the surrounding rock grading level in the construction stage is improved, more comprehensive reference is provided for a subsequent construction scheme, and the construction efficiency is improved. And tunnel intelligent construction is guided.
Owner:SOUTHWEST JIAOTONG UNIV

Urban park physical examination evaluation system construction method based on machine learning algorithm

The invention relates to an urban park physical examination evaluation system construction method based on a machine learning algorithm. The method comprises the following steps: S1) selecting an evaluation index set; s2) taking a park system, an urban comprehensive park, an urban community park and an urban micro-greenbelt collaborative group as evaluation scales, and selecting evaluation indexes from the evaluation index set; s3) acquiring specific data of evaluation indexes to form an index data set; s4) screening repeated indexes with complex collinear features in all index data sets by using a variance expansion coefficient; s5) clustering by using a K-medoids clustering algorithm, and S6) based on an SHAP machine learning algorithm, performing full sample analysis and outputting the importance degree of the index factors, and constructing an urban park physical examination assessment system based on the importance degree of the index factors.According to the method, subjectivity and uncertainty of a traditional assessment method can be avoided, the assessment efficiency is improved, and rapid updating can be performed.
Owner:SHANDONG JIANZHU UNIV

Soil health evaluation method and device based on multiple functions of soil

The invention relates to a soil health evaluation method and device based on soil multifunction, and the method comprises the steps: collecting a soil health index value of a to-be-evaluated region according to a preset collection index; performing function division on the collection indexes to obtain a plurality of soil function categories; respectively calculating soil function scores corresponding to a plurality of soil function categories according to the soil health index values; performing clustering analysis on the plurality of soil function categories of the to-be-evaluated region to obtain weight distribution of the soil function categories; according to the weight distribution and the soil function score, calculating to obtain a soil multifunctional index of the to-be-evaluated region; and performing soil health evaluation on the to-be-evaluated area according to the soil multifunctional index to obtain an evaluation grade. According to the invention, the performance of the soil in different functional aspects can be reflected more accurately, and the soil health condition can be evaluated accurately.
Owner:INST OF AGRI RESOURCES & REGIONAL PLANNING CHINESE ACADEMY OF AGRI SCI

Evaluation and prediction system for water environment bearing capacity of lake basin in arid region

The invention discloses an arid region lake basin water environment bearing capacity evaluation and prediction system, and relates to the technical field of lake drought early warning. A training data collection module is arranged to collect historical meteorological data, historical water level label data, historical meteorological characteristic data and historical drought duration label data; a training sample construction module is set to construct a plurality of lake water level training samples and a plurality of drought duration training samples, a model training module is set to train a water level sequence prediction model and a drought duration prediction model, and a bearing capacity prediction module is set to output future drought duration based on future meteorological characteristic data and the drought duration prediction model. Outputting a future lake water level sequence based on the future meteorological data, the future drought duration, the water level sequence prediction model and the real-time lake water level, and performing early warning on the lake bearing capacity based on the future lake water level sequence; and the prediction accuracy of lake water level change and drought occurrence is improved.
Owner:XINJIANG INST OF ECOLOGY & GEOGRAPHY CHINESE ACAD OF SCI

OCR (optical character recognition) method and system for PCB (printed circuit board) component printing and silk-screen printing

The invention relates to the technical field of printing and silk-screen identification, and particularly discloses an OCR (Optical Character Recognition) method and system for PCB (Printed Circuit Board) component printing and silk-screen. PCB images under visible light, near-infrared and ultraviolet light wavebands are acquired by utilizing a multispectral imaging technology, the visibility of characters is improved, and the interference of a high-reflection or shadow region is inhibited; the quality coefficient of the image is obtained by calculating the structural similarity of the image on different scales, and the quality coefficient is compared with a preset threshold value, so that a high-quality image is automatically screened out; the method comprises the following steps of: comprehensively analyzing each region by extracting contrast, brightness and sharpness data of a high-quality image, and determining and marking effective character regions; a dynamic time warping algorithm is adopted to calculate a deviation coefficient between an actual silk-screen characteristic and a design file characteristic, and whether a printing error exists in silk-screen or not is accurately recognized by comparing the deviation coefficient with a preset threshold value.
Owner:HUAYI INTELLIGENT EQUIP (HANGZHOU) CO LTD

DNA binding protein and RNA binding protein classification method based on ESM-2 and dual-path neural network

ActiveCN119229982Aproportional balancequality improvementBiostatisticsBiological modelsA-DNANucleic acid binding protein
The invention provides a DNA binding protein and RNA binding protein classification method based on ESM-2 and a dual-path neural network, and particularly relates to a staged classification method combining the ESM-2 and the dual-path neural network, which is called DRBP-EDP. In the first stage training, the protein sequence is classified as a nucleic acid binding protein or a non-nucleic acid binding protein; further classifying the nucleic acid binding protein as a DNA binding protein or an RNA binding protein in the second stage of training; meanwhile, a fine data set construction method is further constructed, so that a high-quality protein classification data set is created, through the method, the DRBP-EDP model obtains good performance, the classification accuracy of the nucleic acid binding protein and the non-nucleic acid binding protein in the first stage is 90.03%, and the classification accuracy of the non-nucleic acid binding protein in the second stage is 90.03%. In the second stage, the classification accuracy of the DNA binding protein and the RNA binding protein is 89.56%, a new insight and a new method are provided for protein classification, an innovative tool is provided for research on protein functions, and therefore a new opportunity is provided for development of life science.
Owner:SANYA NATIONAL INSTITUTE OF SOUTHERN BREEDING CHINESE ACADEMY OF AGRICULTURAL SCIENCES +1

Phi-OTDR signal denoising method and system based on improved VMD algorithm, and storage medium

The invention discloses a phi-OTDR signal denoising method and system based on an improved VMD algorithm and a storage medium, and relates to the field of signal processing, and the method comprises the steps: employing a ZOA algorithm to optimize a penalty factor and a decomposition number of the VMD algorithm, and obtaining an optimal penalty factor and a decomposition number; inputting the obtained original phi-OTDR signal into a VMD algorithm which is operated by adopting an optimal penalty factor and a decomposition number to obtain a plurality of IMF components; a deep learning model is adopted to determine a first IMF component needing to participate in signal reconstruction in the IMF components, and the remaining IMF components are to-be-processed IMF components; carrying out denoising processing on the to-be-processed IMF component by adopting an improved wavelet threshold algorithm to obtain a denoised IMF component, and determining a second IMF component suitable for signal reconstruction in the denoised IMF component by adopting a deep learning model; and reconstructing the first IMF component and the second IMF component to obtain a de-noised signal. According to the invention, noise can be suppressed more effectively, and more signal details are reserved, so that the definition and analyzability of the signal are improved.
Owner:JARI AUTOMATION CO LTD CHINA

Water conservancy project leakage abnormity automatic identification method and system based on unsupervised learning

The invention relates to the technical field of engineering geophysical exploration leakage detection, and discloses a water conservancy project leakage abnormity automatic identification method and system based on unsupervised learning, and the method comprises the steps: obtaining detection data, carrying out the preprocessing of the detection data, carrying out the fitting of the detection data through an improved K-means clustering algorithm, obtaining the feature clustering point data of an abnormal region, and carrying out the recognition of the leakage abnormity of a water conservancy project. The method comprises the following steps of: acquiring different leakage abnormal clustering point clusters, correcting the leakage abnormal clustering point clusters, classifying by using an improved DBSCAN algorithm to obtain different leakage abnormal clustering point clusters, communicating the clustering point clusters belonging to the same classification together to form an abnormal clustering point region, aggregating clustering points in the abnormal clustering point region into continuous abnormal region surfaces by using a two-dimensional convex hull algorithm, and then taking each abnormal region surface as a clustering point, according to the method, a three-dimensional abnormal area is formed through classification by using an improved DBSCAN algorithm, and then abnormal area surfaces are aggregated into a leakage channel in a three-dimensional space by using a three-dimensional convex hull algorithm, so that the efficiency can be improved and the deviation caused by artificial experience can be reduced under the condition of ensuring the precision.
Owner:CHINA WATER RESOURCES BEIFANG INVESTIGATION DESIGN & RES CO LTD

Automobile reliability improving method, device and equipment based on FMEA data and medium

The invention relates to an automobile reliability improving method, device and equipment based on FMEA data and a medium, and belongs to the technical field of automobile reliability design, the method is used for early warning risks in the automobile design stage and predicting faults in the automobile operation stage, and the method comprises the steps that automobile faults are layered based on the obtained FMEA data; a fault layering model is generated, and after semantic information extraction is carried out on the FMEA data, a fault database is established; blocks and attributes of the blocks are determined based on the fault database, and a block attribute library associated with the FMEA data is constructed based on the attributes of the blocks; in the design stage, risk early warning is performed on the block scheme in the design stage based on the block attribute library, and in the operation stage, the fault probability prediction is performed on the block scheme by utilizing the fault prediction model which is completely trained to obtain a fault prediction result, so that the reliability of the automobile is improved.
Owner:DONGFENG OFF ROAD VEHICLE CO LTD

Maintenance task distribution method and system based on GBRT algorithm

The invention discloses a maintenance task distribution method and system based on a GBRT algorithm, and belongs to the field of short-term load prediction, and the maintenance task distribution method based on the GBRT algorithm comprises the steps: S1, collecting and preprocessing data; s2, establishing an efficiency model of team work order execution; s3, evaluating the long-term operation effect of the equipment after the maintenance of the team; s4, training and iterating a model of a plurality of decision tree combinations by using a GBRT algorithm; s5, predicting the expected efficiency and effect of task execution of each team by using the trained GBRT model; and S6, generating reasonable maintenance task distribution suggestions including teams to which the tasks are distributed, predicted completion time and required resources. According to the invention, historical data and real-time working information are fully utilized, and the scientificity and accuracy of maintenance task allocation are improved; a task distribution strategy can be adjusted in real time according to current work arrangement and team states, and optimal configuration of resources is ensured; and the overall operation and maintenance efficiency and the equipment reliability are improved through model prediction.
Owner:SANXIA JINSHAJIANG YUNCHUAN HYDROPOWER DEV CO LTD

Artificial intelligence task scheduling method based on deep learning

The invention belongs to the technical field of task scheduling, and provides an artificial intelligence task scheduling method based on deep learning, which comprises the following steps: analyzing and processing a task by using a deep learning algorithm, performing deep learning and processing on the task based on performance data to obtain a task evaluation value, grading the task according to the task evaluation value, and performing task scheduling. The method comprises the following steps: acquiring task data based on comparison of a task evaluation value and a task evaluation standard value, acquiring a task completion value and an advanced task completion value based on the task data, processing to obtain an advanced performance value, comparing the advanced performance value with an advanced performance threshold value, marking a worker according to a comparison result, and marking the worker according to the comparison of the advanced performance value. And comparing the number of the high-level tasks with the number of the high-level personnel to obtain an optimized scheduling value, carrying out combined analysis on the optimized scheduling value YH and the high-level performance value BX to obtain an optimized scheduling coefficient, redistributing the high-level tasks according to the optimized scheduling coefficient, and establishing a redistribution task library.
Owner:SHENZHEN MACROFLASH TECH CO LTD

Diabetes risk assessment method, terminal equipment and storage medium

The invention discloses a diabetes risk assessment method, terminal equipment and a storage medium, multi-modal diabetes data of a user is acquired, and the multi-modal diabetes data comprises clinical diagnosis information, medical images, gene data and lifestyle data; extracting features of the multi-modal diabetes data, and splicing the data features of different modals according to rows or columns to obtain a feature matrix; mining the feature matrix to obtain an association rule; training different N machine learning models by using the screening features, and splicing outputs of the N machine learning models to obtain a new feature vector; and taking the new feature vector as the input of a meta-model, and training the meta-model to obtain an evaluation model. According to the invention, by integrating the multi-modal data, the comprehensiveness and richness of the information are improved, and the risk of diabetes can be evaluated more comprehensively.
Owner:CENT SOUTH UNIV

Power transmission line stockbridge damper defect detection method based on multiple tasks

The invention provides a multi-task-based power transmission line stockbridge damper defect detection method, which comprises the following steps: acquiring an original power transmission line stockbridge damper image, and generating a data sample close to a real fault sample; a multi-task detection network model is established, a classification loss function and a segmentation loss function are constructed for the multi-task detection network model, and the established multi-task detection network model comprises a shared feature layer, classification branches and segmentation branches; based on a data sample, training the multi-task detection network model by adopting an alternate training mode to obtain a trained multi-task detection network model, and in the training process, adaptively adjusting weight parameters by adopting an adaptive dynamic weighting strategy; and inputting a to-be-detected image of the vibration damper of the power transmission line into the trained multi-task detection network model to obtain segmentation positioning and defect detection results of the vibration damper of the power transmission line. According to the invention, the problem of poor robustness and detection performance in the prior art can be solved.
Owner:EAST CHINA JIAOTONG UNIVERSITY

Fine management method for Chinese rose production monitoring based on big data

The invention relates to the technical field of Chinese rose production monitoring, and particularly discloses a fine management method for Chinese rose production monitoring based on big data. According to the invention, the basic parameters of the soil environment of each target area in the planting base corresponding to the current monitoring time period are monitored and analyzed, and then the abnormal values of the soil conditions of each target area in the planting base corresponding to the current monitoring time period are analyzed, so that the defect of real-time monitoring of growth and production of Chinese roses in the prior art is overcome; meanwhile, the reliability and the accuracy of a Chinese rose growth soil environment monitoring analysis result are improved to the maximum extent, reliable data support can be effectively provided for follow-up soil improvement and Chinese rose growth regulation and control work, and the oneness and the subjectivity of Chinese rose growth production monitoring in the prior art are further avoided; and the accuracy of real-time monitoring of Chinese rose production is greatly improved.
Owner:HUAIAN COLLEGE OF INFORMATION TECH

Community energy consumption prediction method, device and equipment based on Bayesian calibration

The invention provides a community energy consumption prediction method, device and equipment based on Bayesian calibration, and relates to the technical field of energy consumption prediction.The method comprises the steps that historical energy consumption data and community energy influence data are input into a trained energy consumption prediction model, first prediction data are generated, and difference data are determined; extracting a time sequence histogram, determining influence data distribution characteristics, and determining prior distribution of the energy consumption prediction model; based on the Bayesian theorem, determining a likelihood value according to the difference data, and generating posterior distribution; inputting the posterior distribution and the future community energy influence data into a trained energy consumption prediction model to generate second prediction data and prediction confidence; extracting a first similarity and a target time interval; respectively extracting historical similar energy consumption data and predicted similar energy consumption data, and determining a second similarity; and generating a final energy consumption prediction result according to the second prediction data, the prediction confidence and the second similarity. According to the invention, the prediction precision can be improved.
Owner:NINGBO ZERO ENVIRONMENTAL SCI-TECH CO LTD

Tunnel water gushing disaster prediction method and device and storage medium

The invention relates to the technical field of tunnel water inrush safety assessment, in particular to a tunnel water inrush disaster prediction method and device and a storage medium, and the method comprises the steps: S1, building a data set based on tunnel water inrush disaster cases, analyzing and selecting disaster-inducing factors, and determining disaster grading standards; s2, dividing the data set into a training set and a test set according to a set proportional relation; s3, performing data preprocessing on a training set and a test set in the data set by taking the disaster-causing factor as a characteristic variable; compared with the prior art, the tunnel water inrush disaster prediction method has the beneficial effects that the tunnel water inrush disaster is intelligently predicted by using the machine learning method, the problems of strong subjectivity, poor applicability and the like of a traditional prediction method can be avoided, meanwhile, disaster misjudgment and missed judgment caused by technicians are avoided, and the model performance and the prediction precision are improved.
Owner:RES INST OF HIGHWAY MINIST OF TRANSPORT +1

Shield parameter geological recognition calculation method and system based on unsupervised clustering algorithm

The invention belongs to the technical field of stratum recognition, and particularly relates to a shield parameter geological recognition calculation method and system based on an unsupervised clustering algorithm, and the method comprises the steps: selecting a plurality of groups of data containing factors affecting stratum category recognition; pCA converts a group of complex related variables into fewer and irrelevant variables, and the new variables are called as main components and can capture most of information in original data; determining the clustering number of K-means + + by adopting an elbow method; determining the clustering number of a hierarchical clustering method by adopting a hierarchical tree; k-means + + and hierarchical clustering are respectively used for clustering the data, and evaluation indexes such as identification precision or contour coefficients of the data and the data are calculated, so that potential modes in the data are analyzed and mined more comprehensively. The invention aims to deeply explore the internal relevance between shield tunnel parameters and geological types, provides an innovative and robust solution for real-time detection of geological features in shield construction by overcoming the limitation of the prior art, and has wide application potential.
Owner:SHENZHEN UNIV

Pedestrian crossing tolerance time prediction method under signal control condition

The invention discloses a pedestrian crossing tolerance time prediction method under a signal control condition. The method comprises the following steps: calculating a risk score of each pedestrian sample in a pedestrian crossing data set through a random survival forest model; carrying out tolerance grouping on the pedestrian samples by using a K-means clustering algorithm in combination with the risk score, and generating a data set of a plurality of tolerance; dividing the data set of each tolerance into a training set and a test set; a Stacking integration model is constructed; performing hyper-parameter tuning on each primary model through a Bayesian hyper-parameter optimization algorithm; inputting the training set into each primary model for training according to five-fold cross validation, and outputting a prediction result of each primary model; integrating the prediction result of each primary model and the true value of the tolerance time into a new training set and a new test set, inputting the new training set and the new test set into the meta-model, and outputting a corresponding tolerance time prediction value; and realizing tolerance time prediction of a single pedestrian through the trained Stacking integrated model. According to the invention, the accuracy of pedestrian crossing tolerance time prediction can be improved.
Owner:CHONGQING JIAOTONG UNIV

Three-dimensional scanning distance and range determination method under given precision condition

The invention relates to the technical field of three-dimensional scanning, in particular to a method for determining a three-dimensional scanning distance and range under a given precision condition, which determines the precision of a three-dimensional scanner within a specified range under different distances through a calibration test and detailed data analysis of a system, provides reliable scanning parameter selection, and improves the accuracy of the three-dimensional scanning distance and range. The method is widely applied to the fields of building measurement, industrial detection, cultural heritage protection and the like; according to the method, subjectivity and blindness of scanning parameter selection are avoided through a scientific parameter selection method, meanwhile, the working efficiency of scanning operation is improved, accidental errors are reduced through multiple tests and data averaging processing, comprehensive data are provided, influence factors and rules of scanning precision are clarified, the scanning process is optimized, and the scanning accuracy is improved. The efficiency and the result precision are improved, and the problem that the precision reliability of an existing three-dimensional scanning system in practical application is limited is solved.
Owner:CHONGQING UNIVERSITY OF SCIENCE AND TECHNOLOGY +2

Economic optimization design method and system for leisure agriculture planning

The invention provides an economic optimization design method and system for leisure agriculture planning, and relates to the technical field of leisure agriculture. The economic optimization design method for leisure agriculture planning comprises the following steps: collecting various data required by leisure agriculture planning, and cleaning and preprocessing the data; performing deep analysis on the preprocessed data, and constructing an economic model of leisure agriculture planning based on the data; in combination with a GIS and an optimization algorithm, performing efficient evaluation on the data, and based on an evaluation result, generating a plurality of preliminary leisure agriculture planning schemes; an interactive decision-making interface is provided, and a decision maker is allowed to adjust planning details according to real-time feedback; and the optimized planning scheme is displayed by a three-dimensional visualization technology, so that a decision maker and an investor can intuitively understand the planning effect. According to the method, the balance between economic benefits and environmental benefits can be realized, the sustainable development capability of leisure agriculture is improved, and the method has important significance for promoting the development of the leisure agriculture.
Owner:INST OF SCI & TECHNICAL INFORMATIONCATAS

Underwater image structure disease full-automatic marking method and system

The invention discloses an underwater image structure disease full-automatic marking method and system, and relates to the technical field of image processing, and the method comprises the steps: obtaining an underwater image of a structure, carrying out the image enhancement, and screening an image with the optimal enhancement effect through employing an image evaluation index; for the image with the optimal enhancement effect, acquiring a disease contour by adopting an OpenCV enhancement method, and drawing a rectangular annotation box; carrying out a clustering method on the obtained disease contour and the rectangular labeling box to obtain a disease label of a part of the image; and performing similarity comparison on the image to be labeled and the image with the obtained label, selecting the label of the image with the highest similarity with the image as the disease label of the image to be labeled, and completing labeling of all the images. According to the method, rapid and accurate marking of underwater image structure diseases is realized, the problems of subjective factors and time and labor consumption of manual marking are avoided, and meanwhile, training does not need to depend on marked label data, so that the whole marking process is simpler, more convenient and more accurate, and better universality and practicability are achieved.
Owner:GUANGDONG LAB OF ARTIFICIAL INTELLIGENCE & DIGITAL ECONOMY (SZ)

Ear canal CT image segmentation method based on improved nn-UNet

The embodiment of the invention discloses an ear canal CT image segmentation method based on improved nn-UNet, and the method comprises the steps: collecting and marking the ear canal CT data of a plurality of patients, and making a data set; preprocessing the data; training on a training set by using a model improved based on nn-UNet, dividing a data set for multiple times by using a five-fold cross validation method, then training and verifying each fold, and storing a training model after training is completed; reasoning and testing the trained model on the test set data which are not seen; post-processing is carried out, and an output result of the model is optimized; and formulating an evaluation scheme to evaluate the prediction result. According to the method, the segmentation efficiency and accuracy can be greatly improved, and subjective deviation and human resource waste possibly generated in the manual segmentation process can be effectively avoided.
Owner:SHANGHAI MARITIME UNIVERSITY

A method and system for automatically detecting surface defects of a flexible touch screen

The present invention discloses a method and system for automatically detecting surface defects of a flexible touch screen, comprising the following steps: S1: using the Retinex method to compensate for the illumination of the surface image of the flexible touch screen, and converting it to the HSV color space to obtain the converted image; S2: calculating the directional feature map based on the converted image; S3: constructing a Gaussian mixture model based on the directional feature map, and clustering the features using the expectation maximization algorithm to obtain the initial defect area; S4: optimizing the initial defect area using morphological processing to obtain the optimized defect area; S5: splicing the converted image and the optimized defect area to obtain an input image, and performing defect detection on the input image based on a multi-scale feature fusion network of deep learning. The present invention can realize fast and accurate detection of surface defects of a flexible touch screen.
Owner:LENGSHUIJIANG JINGKE ELECTRONIC TECH CO LTD

Pattern printing defect detection method for medicine packaging box

The invention discloses a medicine packaging box pattern printing defect detection method, and relates to the technical field of printing quality detection, and the main scheme is as follows: collecting to-be-detected printing data and standard printing packaging box template data, and then calculating a character deviation index, a pattern position coordinate index and a to-be-detected color index according to the data; wherein the character deviation index is obtained by the pixel and format similarity index of the character area; the pattern position coordinate index is calculated based on the standard position coordinate system number and the data to be measured; a color index to be detected is calculated by a deviation color coefficient and the like, then text, position and color index thresholds are set, whether a text area, a pattern position and a pattern color are qualified or not is judged by comparing each index with a corresponding threshold, the printing defect is judged to exist as long as one of the indexes is unqualified, and the method can effectively detect the printing defect.
Owner:ZHUHAI WANJIANG PRINTING CO LTD

Engineering consultation service evaluation method and system

The invention discloses an engineering consultation service evaluation method and system. The method comprises the following steps: firstly, acquiring a plurality of parameters of a plurality of working procedures in the whole process of a project to be evaluated, then acquiring text description of a problem to be rectified of the project to be evaluated, and finally, determining the project to be evaluated according to the text description. The method comprises the steps of obtaining a plurality of parameters of a plurality of processes, carrying out feature interaction analysis on the plurality of parameters of the plurality of processes and the text description of the to-be-rectified problem of the to-be-evaluated project to obtain the semantic interaction feature between a project state and the to-be-rectified problem, and finally determining the estimation duration for the to-be-rectified problem based on the semantic interaction feature between the project state and the to-be-rectified problem. Therefore, the problems of subjectivity and low efficiency caused by manual evaluation can be avoided, more accurate rectification time consumption estimation is provided by utilizing the mutual influence and association relationship between each process link and each factor, and powerful support is provided for project management.
Owner:ZHEJIANG JINGWEI ENG PROJECT MANAGEMENT CO LTD

Concrete appearance quality comprehensive evaluation method and system

The invention discloses a concrete appearance quality comprehensive evaluation method. The method comprises the following steps: S1, collecting concrete appearance image data and carrying out data labeling on a quality defect area; s2, constructing a deep learning instance segmentation network based on the concrete appearance image data obtained in S1, and performing training to obtain a concrete appearance comprehensive defect segmentation model; then, inputting a concrete appearance picture into the concrete appearance comprehensive defect segmentation model, obtaining defect area parameters through image segmentation model reasoning and image analysis, and making a data table; and S3, performing appearance quality scoring on each concrete appearance photo as a target value, and establishing a concrete appearance quality evaluation machine learning model based on the defect area parameters obtained in the S2 as characteristic values, and outputting an evaluation result. Secondary integration is needed, and corresponding evaluation information cannot be given to engineers in time.
Owner:WUHAN YUCHENG QIANLI CONSTR CO LTD

Question and answer method and device, computer equipment and storage medium

The invention belongs to the technical field of artificial intelligence, and relates to a question and answer method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining a target question input by a user, employing a text embedding model to carry out vectorization processing on the target question, and obtaining a question vector; obtaining a fact entry vector set extracted in advance based on the first language model and the text embedding model, and calculating the similarity between the problem vector and each fact entry vector in the fact entry vector set; obtaining at least one fact entry vector with the highest similarity with the problem vector, and determining a fact entry corresponding to the fact entry vector as a target fact entry; and adopting a second language model to generate reply content for the target question according to the target question and the target fact entry. In addition, the invention also relates to a block chain technology, and data such as a target problem and a fact entry vector set can be stored in a block chain. The method can reduce the burden of manually extracting question-answer pairs, and improves the question-answer efficiency and accuracy.
Owner:PING AN BANK CO LTD