624 results about "Rapid identification" patented technology

The invention discloses a wide-area landslide rapid identification method based on an interpretable intelligent algorithm, and relates to the field of remote sensing science and technology, and the method comprises the steps: building a dual-channel feature extraction architecture through multi-source spatio-temporal data fusion and knowledge graph dynamic weighting: capturing image local textures through a lightweight CNN, modeling geological spatial correlation through a graph convolutional network, and carrying out the recognition of the landslide. Combining the SHAP value and causal reasoning to generate an interpretable contribution degree thermodynamic diagram and a rule chain; a knowledge graph bidirectional verification system is introduced, spatial logic contradictions are verified by using prior rules, and co-evolution of a model and a rule base is triggered based on misjudgment samples; outputting a multi-dimensional credibility report, quantifying uncertainty by Monte Carlo Dropout, and customizing interpretation granularity according to roles; a terrain-adaptive block-stream processing architecture is adopted, and edge lightweight deployment and federated learning are combined, so that wide-area real-time early warning and model dynamic updating are realized. According to the scheme, the limitation of a traditional black box model is broken through, and a disaster prevention closed loop with physical driving, transparent decision and second-level response is formed.

The invention discloses a system and method for visually detecting the quality of a pipe pile. The system comprises a bearing module used for bearing the pipe pile, an end detection module and a movable pipe pile body scanning module. The end detection module is used for collecting end face data of the end of the pipe pile and recognizing the defects of the end of the pipe pile according to the end face data of the end of the pipe pile. The movable pipe pile body scanning module moves along a pipe pile and is used for collecting pipe pile body data, and pipe pile body defects are recognized through the pipe pile body data. And summarizing the pipe pile end defects and the pipe pile body defects to generate a defect report. Rapid identification, quantitative analysis and data archiving of pipe pile defects are achieved, and the detection precision and the production efficiency are remarkably improved.

The invention relates to the technical field of environment monitoring, in particular to an intelligent water quality monitoring method and system for collaborative sampling of a bionic fish school and an unmanned aerial vehicle and computer equipment. The method comprises the following steps: dividing monitoring grid units by using GIS data, establishing a layered communication architecture, dynamically deploying a bionic fish school to carry out real-time water quality monitoring, and scheduling and optimizing resource allocation through a fuzzy rule; historical pollution event data and real-time sensing data are integrated to construct a water quality evaluation index system, and a Nash equilibrium strategy is adopted to determine an optimal weight to generate a dynamic pollution situation cloud chart; constructing an inferior fingerprint database to realize rapid identification of pollution events and trigger a two-stage response mechanism; and finally, combining the bionic fish school sampling data and the unmanned aerial vehicle pre-diffusion coordinate to dynamically simulate the effect of the treatment measure. Air-water integrated efficient monitoring is achieved, the accuracy and timeliness of water quality anomaly detection are improved, and comprehensive technical support is provided for water area environment protection.

The invention provides a portable detection device and method for rapidly identifying peritonitis, and relates to the technical field of medical instruments, the device comprises: an illumination module for irradiating a peritoneal dialysis liquid sample; the sensor module is used for capturing a sensing data set of the state change of the peritoneal dialysis liquid sample in the irradiation process; the detection chip downloads the trained peritonitis risk detection model, performs data detection on the sensing data set and outputs a risk detection result; and the display module is used for displaying on a display screen. The peritonitis risk detection device can solve the technical problem that a user is difficult to independently complete rapid and accurate detection in a family or community environment due to poor portability and complex operation of a traditional peritonitis detection device, and achieves the purposes of rapidly analyzing a peritonitis sample through optical characteristics and a machine learning algorithm, outputting a peritonitis risk detection result in real time and improving the detection efficiency. And the accuracy and the efficiency of early peritonitis recognition of the user or the medical personnel are improved.

The invention belongs to the technical field of precise and intelligent breeding, and provides a sheep apparent state recognition method and system based on an improved supervised learning network, and the method comprises the steps: apparent state category determination, data collection, dynamic trajectory prediction and filtering noise reduction, superposition windowing processing and feature extraction, data merging, data set grouping and model training. According to the method, dynamic trajectory prediction and filtering noise reduction are carried out on the collected data, and superposition windowing processing and feature extraction are carried out on the preprocessed data, so that the noise of the training data is reduced, the identification precision of the model is improved, scientific, accurate and rapid identification of the sheep apparent state is realized, and the identification efficiency of the sheep apparent state is improved. And breeding personnel can be helped to know the physiological and health conditions of the sheep in time.

The invention discloses a rapid identification and comprehensive evaluation method for a polygonatum cyrtonema provenance, and belongs to the field of improved variety breeding and quality control of traditional Chinese medicinal materials. The method systematically measures a plurality of specific biological traits covering three dimensions of growth, physiology and medicine. And carrying out standardization processing on the measured multi-dimensional data, carrying out dimensionality reduction by using a principal component analysis model, constructing an evaluation model capable of comprehensively reflecting the comprehensive quality of the provenance, and calculating a quantitative comprehensive evaluation score. The score can objectively and scientifically divide different provenances into four different quality grades including excellent, good, common and to-be-improved quality grades. The method has the beneficial effects that the strong positive correlation between the easily measured growth traits of the overground part and the medicinal component content of the underground rhizome for determining the final quality of the medicinal material is established, so that the rapid, predictive and non-destructive early screening of the provenance quality is realized.

The invention provides a multi-mode underwater biological intrusion detection method and device, and the method comprises the steps: carrying out the sampling of each sampling grid of a target water area through acoustic equipment and image equipment, and obtaining the acoustic data and image data of each sampling grid; extracting and fusing the acoustic data and the image data of each sampling grid to obtain multi-modal features of each sampling grid, taking the multi-modal features as input of a discriminator in the generative adversarial network, and obtaining a discrimination result output by the discriminator as existence of invasive organisms or absence of the invasive organisms in the sampling grids; under the condition that the invasive organism exists in the sampling grid, sampling and detecting the eDNA environment sample on the sampling grid, and determining whether the invasive organism exists in the sampling grid or not based on a detection result. According to the method, rapid identification is carried out by utilizing multi-modal information and acousto-optic characteristics, the identification accuracy is improved by utilizing eDNA, the defects of a single method are overcome, and the efficiency of monitoring invasive species is improved.

The invention discloses a rainfall intensity rapid identification method and system based on multi-sensor data fusion, and relates to the technical field of hydro-meteorological monitoring and urban drainage scheduling, and the method comprises the steps: obtaining a historical rainfall data set, drawing a peak curve, and extracting an evolution sequence feature before a peak value; correlation analysis is carried out on the data and complete rainfall process characteristics to establish a historical rule model; rainfall intensity data are collected in real time through multiple sensors, and space-time dimension features are fused; matching the fusion features with historical features, and screening target historical events; constructing a rainfall reproduction prediction model to reconstruct a current rainfall evolution process; and finally, calculating an inflow load based on a reconstruction result, and generating a dynamic drainage scheduling scheme. According to the invention, accurate prediction of the rainfall process and intelligent scheduling of the drainage system are realized, and the urban waterlogging prevention capability is improved.

The invention relates to the technical field of dynamic monitoring, in particular to a chemical production safety risk dynamic monitoring system based on the Internet of Things, and the system comprises a data collection module which receives pressure monitoring data, temperature fluctuation data and pipeline flow velocity data measured by a chemical plant sensor in real time, records the time of each data point, and transmits the data to a server; and sorting and organizing the data according to a time sequence. According to the method, by receiving and sorting the chemical plant sensor data and implementing comprehensive analysis of the multi-scale time window, the capture and analysis precision of key parameter fluctuation is remarkably improved, the scientificity of rapid identification and risk prediction of variable fluctuation in the chemical process is promoted, potential risks can be found in advance, and the risk prediction efficiency is improved. Therefore, production strategies are adjusted in real time, production safety is ensured, pre-judgment of future risks is optimized by analyzing long-term trends and periodic changes, the accuracy of risk assessment and the safety of operation are improved, the probability of accidents is reduced, and the safety of personnel and equipment is protected.

The invention discloses a fault identification processing method and device based on an internet of things digital weighing system, and the method comprises the steps: a fault detection step: tracking and recording the weighing data and state of each sensor of the internet of things digital weighing system, and judging whether the sensors have faults or not; and an output replacement step: if a sensor fails, generating predicted weighing data according to the historical weighing data of the sensor and the weighing barycentric coordinates of the weighing platform, taking the predicted weighing data as the output of the failed sensor, and outputting the weighing value of the whole scale in combination with the weighing data of other normal sensors. According to the invention, rapid identification and early warning can be realized, and the influence of abnormal data of the sensor can be effectively eliminated or reduced by learning and utilizing historical weighing data and a weighing gravity center change rule of the weighing platform, so that the reliability of the internet-of-things digital weighing system can be effectively improved; and production losses such as shutdown and production halt caused by the abnormal problem of the sensor of the weighing system of the field user are reduced.

The invention discloses a rapid diagnosis method for blueberry continuous cropping soil obstacle factors, and belongs to the technical field of soil environment analysis. Based on combined analysis of soil physical and chemical factors, metagenome and metabolome, pathogenic bacteria types and key metabolites significantly related to the blueberry continuous cropping obstacle are extracted, and a characteristic diagnostic index system is formed. The index system comprises main pathogenic bacteria types, typical harmful metabolites, key soil physical and chemical factors and appropriate or risk threshold ranges thereof, and provides a scientific basis for rapid identification of soil continuous cropping obstacles. Comparative analysis is carried out by establishing a standardized index library, rapid comparison of related indexes and standard indexes in a newly sampled soil sample is realized, and whether the soil has continuous cropping obstacle risks or not is judged in an auxiliary manner by integrating the deviation degree of physical and chemical factors and biological factors.

The invention provides a tunnel back break intelligent rapid identification technology based on machine vision. According to the method, firstly, the actual space distance of a reference point on a tunnel face is obtained through a laser range finder, and the pixel and physical size conversion proportion is calculated in combination with a shot tunnel face image; and then, screening high-quality guide information by using a visual language model CLIP, and driving a segmentation model SAM for fine adjustment of a tunnel scene to realize accurate segmentation of the tunnel face contour. The contour pixel area obtained through segmentation is converted into the actual physical area in combination with the conversion proportion, the actual physical area is compared with the preset tunnel design contour area, and therefore the over-excavation or under-excavation state of the tunnel face is automatically judged, and the size of the over-excavation or under-excavation area is determined. According to the method, the automation degree, precision and real-time performance of back break detection in tunnel engineering are remarkably improved, the limitation of traditional manual measurement is overcome, and key technical support is provided for intelligent tunneling of a tunnel.

The invention provides an unmanned aerial vehicle photovoltaic real-time inspection system and method based on an end-edge-cloud architecture, and belongs to the field of photovoltaic power station operation and maintenance. Comprising an end device which directly interacts with an environment or a user and collects data and comprises an unmanned aerial vehicle end, a flight controller, a pan-tilt camera end and a mobile phone end; the edge computing platform is realized by adopting an embedded platform carried above the unmanned aerial vehicle end, and is used for reading end equipment data, executing fuzzy judgment based on a visible light image and two-stage defect detection based on an infrared image, and uploading an image detection result to the cloud server; and the cloud server is responsible for data storage and message pushing, receiving the image result uploaded by the embedded platform and pushing the image result to the mobile phone terminal. According to the invention, an end-edge-cloud collaborative architecture facing a power station is provided, the diagnosis and decision delay is greatly shortened, the rapid identification and positioning of the defects of the photovoltaic equipment are realized, and the high-precision identification of the defects is realized by adopting a two-stage detection algorithm based on target identification and defect classification.

The invention relates to the technical field of plant genetic engineering, and discloses a haplotype identification marker of a rice flood-resistant gene ARF9 and application of the haplotype identification marker. The molecular marker is derived from an exon of an ARF9 gene, is located at the 22009082bp and 22010019bp of a No.4 chromosome of rice, and has a basic group of C. The SNP molecular marker is related to the waterflooding resistance of a male parent oryza longistaminate of perennial rice, and the waterflooding resistance of the male parent oryza longistaminate with a locus genotype of CC is obviously higher than that of rice with a locus genotype of non-AA type. Therefore, the invention provides the loop-mediated isothermal amplification primer group for rapidly identifying the ARF9 gene haploid molecular marker, and the loop-mediated isothermal amplification primer group is used for identifying the ARF9 gene haploid molecular marker so as to rapidly screen out waterflooding-resistant rice. The invention provides a group of primers, the flood-resistant rice can be quickly identified by adopting the primers to perform loop isothermal reaction, the detection cost is low, the detection speed is high, the flood-resistant germplasm can be quickly screened in the rice seedling stage by adopting the method provided by the invention, a time-consuming flooding test does not need to be performed in the later stage, and the breeding process is greatly accelerated.

The invention relates to the technical field of cardiovascular disease diagnosis, and discloses a cardiovascular disease parameter identification method and system based on data analysis, and the method comprises the steps: collecting cardiovascular image data, physiological parameter data and clinical data of a patient, obtaining multi-source data, carrying out the noise reduction and missing value and abnormal value processing of the multi-source data, and obtaining preprocessed data; calculating a mutual information value between each feature and the cardiovascular lesion by adopting a feature selection algorithm based on mutual information to obtain a target feature; encoding the target features, inputting the encoded target features into a parameter identification model, and processing the encoded target features by the parameter identification model to obtain a cardiovascular lesion parameter identification result; according to the invention, accurate and rapid identification of cardiovascular disease parameters is realized, and the accuracy and efficiency of cardiovascular disease diagnosis are improved.

The invention relates to the technical field of soil cadmium remediation, in particular to a soil Cd recognition and remediation method based on mulberry root exudates and microorganisms. According to the method, spatial microscopic image frames of rhizosphere microflora are divided through a community structure, community node increment updating calculation of adjacent time sequences is executed, the transition recombination frequency of the rhizosphere microflora is obtained, and the transition color of the actual concentration content of root exudates is rendered based on a metal pollution level system and the transition recombination frequency. Generating a stress level heat model of the soil cadmium gradient in the target restoration area; and according to the stress level popularity model, carrying out popularity color analysis and identification on the soil cadmium pollution of the target restoration area, and generating a global stress identification result of the soil cadmium. According to the method, rapid identification of cadmium pollution enrichment can be realized by utilizing the synergistic response between mulberry root exudates and rhizosphere microorganisms, and a repair regulation and control plan is made, so that the cadmium pollution degree is reduced, the soil structure and fertility are optimized, and a safer and healthier environment is provided for plant growth.

The invention belongs to the technical field of bacterium identification, and particularly relates to a method and system for rapidly identifying mixed bacteria in a water body based on a transmission spectrum, and the method comprises the steps: measuring the multi-wavelength transmission spectrum characteristics of a bacteria mixed solution at the wave band of 210-800 nm, building a plurality of identification models through a support vector machine, and employing a layer-by-layer identification strategy to identify the mixed bacteria in the water body. And the mixed bacteria in the water body can be quickly and accurately identified. The method comprises the following steps: preparing a bacterial mixed solution with equal concentration and different volume ratios, measuring a transmission spectrum of the bacterial mixed solution, carrying out minimum-maximum normalization processing, constructing a plurality of identification models for training, and finally carrying out bacterial type identification on a bacterial solution sample to be detected. According to the method, different types of bacterial mixed solutions with the same concentration and different volume ratios are prepared, the multi-wavelength transmission spectrum characteristics of the bacterial mixed solutions are measured, a plurality of recognition models are established, and a layer-by-layer recognition strategy is adopted, so that the recognition accuracy and robustness are effectively improved.

The invention relates to the technical field of biological outbreak prediction, in particular to a marine organism outbreak prediction method and system based on real-time monitoring, and the method comprises the following steps: based on a sea area microsensor array, collecting environment and plankton data, processing interference and noise, normalizing parameters, and combining ecological baseline adjustment; and analyzing the environment and biological response relationship between the nodes, identifying a key area and a change trend, and judging plankton extension characteristics. According to the method, synchronous linkage of environmental factors and plankton response is realized through a joint discrimination mechanism of real-time acquisition, multi-parameter dynamic normalization and fusion of spatial distribution and time sequence data, sea area key change signals are identified by depending on ecological baseline reference, abnormal response is automatically screened in a monitoring network, and an extension situation is tracked; the space-time evolution trend of a risk area is effectively reflected, rapid identification and prospective early warning of potential explosion phenomena are realized through whole-process dynamic evaluation, and a reliable data basis is provided for ecological intervention and scientific management.

The invention discloses application of a KASP molecular marker in identification of sweet melon peel color traits, and belongs to the technical field of molecular markers. The invention aims to promote gene localization and molecular-assisted breeding processes by utilizing molecular marker-assisted selective breeding on sweet melon pericarp color characters. The KASP molecular marker is a KASP marker RW2; the KASP marker RW2 is designed according to an SNP (Single Nucleotide Polymorphism) site at the position of Chr04: 526653. According to the application, the KASP molecular marker RW2 is used for performing genetic typing on the red-peel muskmelon and the white-peel muskmelon, the DNA of a target plant can be identified only through one-time PCR amplification, the whole detection process is simple and convenient, and complex steps such as enzyme digestion, electrophoresis and sequencing are not needed. Through the high-throughput genotyping system, a genotyping graph and a genotyping value can be rapidly obtained, so that rapid identification of the genotype of a target plant is realized. The KASP molecular marker RW2 provided by the invention can be more efficiently applied to the breeding work of sweet melon peel color germplasm.

The invention relates to the technical field of application computer science and artificial intelligence, and discloses a highway pavement disease rapid identification and diagnosis system based on deep learning, and the system comprises a data collection and preprocessing module which is used for obtaining and processing multi-source time series data; the unified state vector characterization module is used for constructing a unified vector for characterizing the comprehensive state of the pavement; the spatio-temporal evolution and inference module is used for simulating the spatio-temporal evolution process of the pavement state; the model training and optimizing module is used for carrying out optimization training on the evolution model; and the diagnosis and prediction application module is used for generating a final diagnosis report and maintenance decision. According to the method, the development trend of the pavement diseases and the effects under different maintenance scenes are deduced and simulated through the spatio-temporal depth model jointly constrained and optimized by physical and causal knowledge, a complete closed loop of the pavement diseases from static recognition to dynamic diagnosis and decision making is realized, and the scientificity and the intelligent level of road maintenance management are improved.

The invention discloses a method and a device for rapidly identifying and tracing wetland disturbance based on multi-modal data. The method comprises the following steps: constructing a multi-modal data acquisition network; performing feature processing on the acquired data information to obtain multi-modal fusion data; constructing a wetland disturbance identification index system; a quantitative evaluation result of the wetland health indexes is obtained; obtaining an index threshold value and an index grade of the target index item; and executing an exception handling operation corresponding to the early warning level. According to the invention, a wetland disturbance identification index system is constructed, and the collected data information is subjected to feature processing to obtain multi-modal fusion data, so that the data utilization efficiency is greatly improved; based on a Bayesian network model, accurate identification and influence degree quantification of an interference source are realized, a proper threshold determination method is selected according to data features, a dual early warning mechanism combining active early warning and response early warning is constructed, and the early warning accuracy is improved through cross validation.

The invention discloses a forest fire rapid identification and accurate fire point positioning method based on air-ground cooperation. Comprising the following steps: a forest fire panoramic monitoring and positioning method based on a double-axis holder, a wildfire target intelligent identification algorithm based on multi-modal image fusion analysis, and a low-power-consumption air-ground cooperative communication network oriented to forest fire monitoring. The invention relates to an unmanned aerial vehicle-based forest wildfire target maneuvering inspection and positioning method and an air-ground fusion wildfire target accurate positioning method. According to the invention, 360-degree panoramic image acquisition is realized by using the double-shaft holder, and meanwhile, a fire source position self-adaptive adjustment mechanism is provided; multi-modal image fusion analysis is adopted in the image processing layer to resist interference, blind areas are compensated, and the fire source recognition accuracy is improved; a low-power-consumption air-ground cooperative communication network is constructed, so that the cost is reduced, and flight is continued; and the monitoring limitation is broken through by means of flexible inspection of the unmanned aerial vehicle, a closed-loop system of air-ground cooperation-information complementation-intelligent verification is formed, and the monitoring stability in a complex scene is remarkably enhanced.

According to the method for rapidly identifying the klebsiella pneumoniae high-viscosity strain, the problems that the detection process is tedious, consumed time is long and the like are solved in the mode that the surface enhanced Raman spectrum is prepared and the machine learning algorithm is combined. And on the other hand, a Raman spectrum database of different klebsiella pneumoniae high-viscosity strains is established through machine learning, newly collected Raman spectrums are directly imported into the database for judgment, and the accuracy of an identification result can be better guaranteed through comparison and combination of different machine learning methods and evaluation indexes.

The invention relates to the technical field of biology, and particularly discloses a kit for detecting haemophilus influenzae based on PCR (Polymerase Chain Reaction) / RAA-CRISPR-Cas13a. The invention discloses a kit for detecting a haemophilus influenzae gene on the basis of PCR (Polymerase Chain Reaction) / RAA-CRISPR-Cas13a. The kit comprises a primer pair for specifically amplifying haemophilus influenzae genes and crRNA, the sequence of the crRNA is composed of an anchoring sequence used for being combined with Cas13a protein and a guide sequence of a target sequence of the haemophilus influenzae genes, and the guide sequence is shown as the 39-66th site of SEQ ID No.28. The invention constructs a method for detecting haemophilus influenzae genes by combining a PCR or RAA technology with a CRISPR-Cas13a technology, the method is simple, convenient and rapid, has higher sensitivity and specificity, can reach a single copy level in sensitivity, and can be used for clinically and rapidly identifying whether the haemophilus influenzae is infected or not.

The invention relates to a gene ZmILR1 for controlling the size of a corn kernel as well as a molecular identification method and application of the gene ZmILR1, and belongs to the technical field of plant genetic breeding and functional gene research. The ZmILR1 gene encodes an IAA-amino acid hydrolase, mutation of the IAA-amino acid hydrolase causes corn kernels to become smaller, and overexpression of the gene can create a corn material with increased kernel length, kernel width and hundred-kernel weight. A functional molecular marker A1A4 is developed according to the sequence difference of the mutant gene and a wild type gene at the DNA level, and can be used for rapid identification and screening of small-grain corn materials. The invention provides a theoretical basis and an application tool for analyzing a molecular mechanism of the corn grain size and developing molecular marker-assisted breeding and hybrid seed breeding, and provides important gene resources and technical support for high-yield and high-quality molecular breeding of corn.

The invention discloses a rapid pathogen identification method and system, and belongs to the technical field of medical information processing. According to the method, space-time growth discrimination operation, multi-scale structure perception operation and microscopic focusing classification operation are combined, the dynamic trend of bacterial colonies from appearing to diffusion is accurately captured, low-contrast textures of micron-scale initial spots and the complex structure of millimeter-scale mature bacterial colonies are considered, a refined space context correlation model is established, and the dynamic trend of the bacterial colonies from appearing to diffusion is obtained. The method solves the problems that a traditional single-frame analysis method is insufficient in continuous modeling of the growth process and early-stage features are confused, the detection sensitivity of early-stage minute colonies is improved, rapid and accurate identification of pathogenic bacteria types is achieved, and the identification requirements in clinical rapid detection and complex microbial environments are met.

The invention discloses a chicken sex identification method based on fluorescence labeling, and belongs to the technical field of gene engineering. The chicken sex identification method comprises the following steps: co-transfecting a CRISPR / Cas9 system and a WPG-EGFP donor plasmid into chicken primordial germ cells PGCs, screening to obtain EGFP positive PGCs, micro-injecting the EGFP positive PGCs into a receptor chicken embryo, hatching to obtain an EGFP reproductive chimera chicken, and identifying the sex through fluorescence labeling, the CRISPR / Cas9 system comprises sgRNA of a targeted chicken W chromosome WPG gene; the WPG-EGFP donor plasmid is constructed by inserting an EGFP gene into a WPG gene locus through a homologous recombination strategy. According to the method, the sex of the chicken can be visually and rapidly identified. The method solves the problems that a traditional sex identification method is complex in operation and depends on sample ages, has the advantages of high efficiency and accuracy, and is suitable for sex control and screening of poultry such as laying hens and broilers.

The invention discloses a method for rapidly identifying trace antibiotics in milk, which solves the problems of tedious detection process and long consumed time by combining preparation of a surface enhanced Raman spectrum and a machine learning algorithm. On the other hand, Raman spectrum libraries of different human milk antibiotics are established through machine learning, newly collected fingerprints are directly imported into the libraries for judgment, and the accuracy of identification results can be better guaranteed through comparison and combination of different machine learning methods and evaluation indexes.

The invention provides a building group structure type identification method and device based on a multi-modal large model, and relates to the technical field of multi-modal large models. The method comprises the following steps: extracting observable building features through a multi-modal large model; establishing a graph enhanced multi-modal large model to perform matching and reasoning on the features and the knowledge graph; an expert auditing mechanism is introduced to correct an identification result, the knowledge graph is dynamically updated, and the graph enhanced multi-mode large model is finely adjusted; and establishing an urban building group rapid identification system, and establishing a rapid identification model based on deep learning and an interpretable identification system based on a graph enhanced multi-modal large model by using a knowledge graph. According to the method, the explainable reasoning capability of deep learning rapid classification and map enhancement is fused, and a structure identification closed-loop system with self-learning and self-updating capabilities is constructed. The method is suitable for scenes such as urban building census, anti-seismic evaluation and BIM auxiliary modeling, and has high accuracy, strong generalization and sustainable evolution capability.

The invention discloses a light-weight Transform tomato disease identification method fused with self-supervision enhancement, and the method comprises the steps: carrying out the data preprocessing of an image through collecting and obtaining the image of a tomato disease; an SAEFormer model is proposed, a multi-scale selection fusion attention module is introduced, and the feature extraction capacity of the model is enhanced; self-supervised learning loss is introduced, so that the modeling capability of the model on fine-grained difference is effectively improved; a RepBN optimization strategy is adopted, so that the model deployment efficiency is improved, and the calculation overhead is reduced; inputting the preprocessed image data into the model for training and verification to obtain an optimal model; and finally, inputting a to-be-classified image into the optimal model to obtain a classification result. According to the light-weight Transform tomato disease identification method fused with self-supervision enhancement provided by the invention, rapid identification of tomato diseases can be ensured, correct types can be obtained, and the method has a good application prospect.