539 results about "Structure recognition" patented technology

The invention discloses an AI and BIM-based project cost control method and system, and the method comprises the steps: obtaining the three-dimensional model data of a target building, analyzing the file structure of the three-dimensional model data, recognizing the type, size parameters and material attributes of a component, building a mapping relation between a component coding system and an attribute tag if the three-dimensional model data passes the integrity inspection, and carrying out the construction cost control of the target building. Generating a standardized component data set; constructing a material price fluctuation prediction model according to the historical price record, intercepting time series data by adopting a sliding window mechanism, and if the price fluctuation amplitude in the material price fluctuation prediction model exceeds a preset threshold, triggering an early warning identifier, and generating cost prediction data with a risk level; and carrying out association mapping on the standardized component data set and a quota library by adopting a coding matching mechanism, and if component attributes are successfully matched with quota entries, converting and generating engineering quantity data based on geometric parameters, and calculating the cost of a single project. The cost prediction accuracy is improved.

The invention relates to the technical field of intelligent manufacturing, and discloses a target structure automatic detection method, device and equipment and a medium, and the method comprises the steps: obtaining scanning path planning data of a target detection structure, driving an ultrasonic probe to execute surrounding scanning motion, and collecting an ultrasonic image sequence and spatial pose data; dynamically adjusting a pressure application angle and a scanning speed based on force feedback information, fusing spatial pose data and an image sequence to perform three-dimensional reconstruction, constructing a three-dimensional geometric model of a target detection structure, extracting feature distribution data by applying an intelligent analysis model, generating a feature decision set, and performing feature extraction; and mapping the feature decision set to a three-dimensional coordinate system to construct an analysis report containing the feature type marks and the topological relation. Through fusion of force control scanning, image reconstruction and intelligent analysis, standardization and intelligence of a detection process are realized, image consistency and structure identification precision are improved, manual dependence is reduced, and comprehensiveness and reliability of lesion identification are enhanced.

The invention provides a semantic enhancement adaptive partitioning method and system for natural resource large model questions and answers, and aims to solve the problems of difficulty in term boundary recognition, damage to semantic integrity and the like. According to the method, three core technologies including theme perception coarse-grained paragraph division, self-adaptive sliding window theme hierarchy division and embedded perception context self-adaptive text segmentation are fused. The system firstly analyzes a natural resource long text structure, identifies titles and theme levels and aligns associated contents; paragraphs are extracted according to a theme perception strategy and are subdivided into sentence sets according to grammar rules; an improved sliding window mechanism is adopted to divide sentences into window sentence block groups. The method is characterized in that a dynamic aggregation threshold mechanism is introduced, the semantic association degree between adjacent sentence blocks is calculated through an embedded perception context semantic segmentation technology, whether the sentence blocks are combined or not is judged by combining a similarity distribution change trend and a dynamic adjustment threshold, self-adaptive delimitation of semantic boundaries is achieved, and text blocks which are clear in structure and coherent in semantics are generated.

The invention discloses a multi-signal quality monitoring system and method in the numerical control machining process, and belongs to the technical field of numerical control machining. The method comprises the following steps: step 1, synchronously acquiring multi-source signals in a microscopic level, a mesoscopic level and a macroscopic level in a processing process; 2, screening to form an optimal feature subset, and constructing a multi-dimensional feature representation space; 3, constructing a same-scale transverse incidence matrix, a cross-scale longitudinal incidence matrix and an interaction incidence matrix based on a bidirectional attention mechanism, and forming an incidence network structure for describing feature coupling and dependency relationships among signals of different levels; step 4, calculating deviation metric values of various signal features, identifying an abnormal diffusion path in combination with a topological structure of the association network, and analyzing a causal relationship chain between the signals; and step 5, matching the identified abnormal mode with a pre-constructed process knowledge graph, automatically generating a targeted process parameter optimization suggestion, and performing real-time evaluation to realize closed-loop adaptive control for quality abnormity.

The invention discloses a box-type transformer substation monitoring method based on unsupervised learning, and the method comprises the following steps: S1, collecting multi-source sensing data in the operation process of a box-type transformer, and generating a standardized input data set; s2, selecting a data sample in a normal operation state, constructing an unsupervised learning model, and obtaining a feature representation set; s3, inputting data, extracting current operation state characteristics, and recognizing a voiceprint abnormal state by combining outlier detection; s4, performing window sliding and statistical analysis on the time sequence data, and outputting a trend abnormal interval and an abnormal index type; s5, constructing a variable association graph structure, and identifying potential abnormal variables and propagation paths; s6, integrating the various types of abnormal information and the feature representation set, and constructing a voiceprint map library; and S7, integrating and processing the abnormal result and the voiceprint map database, and executing visual display and intelligent early warning. According to the invention, box transformer substation abnormity monitoring and early warning based on unsupervised learning are realized, and the fault identification accuracy and response efficiency are improved.

The invention relates to the technical field of artificial intelligence, in particular to a large model and multi-agent collaborative decision-making method based on dynamic knowledge flow, which comprises the following steps of: analyzing a static knowledge and dynamic information fusion relationship through joint modeling, extracting a hierarchical structure of equipment constraints and environment variables, identifying constraint conflicts and deviations in task decomposition, and obtaining a multi-agent collaborative decision-making result; screening a consistency decomposition direction, extracting a task constraint parallel optimization theory, correcting constraint conflicts, evaluating consistency changes, and outputting a collaborative task convergence robust state identifier. According to the method, by integrating static knowledge and dynamic information and optimizing understanding of task decomposition and equipment constraints, the collaborative decision-making capacity of multiple agents in a complex environment is enhanced, the conflict and deviation processing capacity in the task execution process is improved, the accuracy and consistency of tasks are enhanced, and the convergence and stability of task targets are improved; the robustness of multi-agent cooperative work is promoted, and finally more efficient resource utilization and task completion effects are achieved.

The invention relates to the technical field of English language ability assessment, in particular to an English language ability assessment method and system. The method comprises the following steps: collecting a writing text uploaded by a student, and analyzing the literature and style category and sentence pattern structure of the writing text to obtain a content literature and style category sentence pattern structure; thirdly, identifying logic relation deviation in the text based on the structures, performing clustering, further quantizing the context fuzzy degree, and generating context fuzzy degree quantized data; and finally, designing an English language ability evaluation architecture by adopting a Q-learning algorithm. According to the invention, the English language ability evaluation technology is optimized, so that the English language ability evaluation technology is more accurate.

The invention provides a full-text retrieval method and system fusing various types of documents, and relates to the technical field of information retrieval, and the method comprises the following steps: obtaining document representation through document content extraction and structure recognition, generating a cross-modal semantic vector by using word embedding and nonlinear transformation, constructing a hierarchical index and a cross-document association graph, and obtaining a full-text retrieval result; the basic correlation score is calculated after the query request is received, and the comprehensive score of the candidate content segments is calculated based on the association graph to determine the optimal retrieval result, so that unified representation and retrieval of heterogeneous documents are realized, the cross-document retrieval precision and relevance are improved, and the processing capability of a retrieval system on complex queries is enhanced.

The invention provides an OCR character recognition method and system based on an end-to-end network, and relates to the technical field of image recognition, and the method comprises the steps: inputting a to-be-recognized image into a preset neural network for degeneration type recognition and corresponding processing, and obtaining a preprocessed image; performing character recognition and structure recognition on the preprocessed image by using a pre-trained end-to-end comprehensive model to obtain character text information and character structure information; recognizing a character image according to the character text information, and generating a character structure constraint graph based on the character image; and constructing a feedback result based on the character structure constraint graph and the character structure information, and optimizing the end-to-end comprehensive model by using the feedback result. According to the method, the identification precision can be improved, the robustness and the adaptive capability are improved, and the problems that in the prior art, in the face of an actual complex degraded scene, the adaptive capability is lacked, errors cannot be effectively fed back and optimized, and the identification accuracy and robustness are reduced are solved.

Disclosed herein are methods and systems to generate code for a workflow. A non-limiting example of a method comprises executing, by a processor, a large language model, which receives an input of an intent associated with a workflow and provides a machine-readable description by: identifying, by searching a hierarchical data structure using a vector embedding associated with the intent, an object corresponding to the intent; identifying a set of data paths within the hierarchical data structure to retrieve the identified object; and generating the machine-readable description, the machine-readable description describing the object and at least one data path within the set of data paths; and generating, by the processor, code for the workflow using the machine-readable description.

The invention discloses a drilling rock core image automatic identification and recording method and system based on deep learning. The method comprises the following steps: preprocessing a collected to-be-detected drilling rock core image; the preprocessed image is input into a trained Transform semantic segmentation model, a pixel-level probability graph is output, and a binary mask image is generated through binarization processing; performing morphological processing, connected domain analysis and contour detection on the binary mask image to obtain edge coordinate information of the rock core frame, extracting a standardized rock core image from the original image according to the edge coordinate information, and performing result storage and visualization; inputting the standardized rock core image into a trained Transform multi-category rock core structure recognition model, outputting a pixel-level probability graph of multiple channels, and generating a binary mask image of each channel through binarization processing; and three-rate calculation and multi-structure statistical analysis of the drilled rock core are carried out, and data output and visualization are carried out.

The invention discloses an AI-based archive digitization full-process management and control method, which comprises the following steps: acquiring archive image data and text data, constructing an original archive data set, and generating an initial directory set; constructing an archive directory template set, executing template matching and structure adaptation, and generating an initial structure recognition result; setting a classification rule of directory deviation and structural contradiction, constructing a deviation rule base, executing rule scanning and node comparison, and generating a deviation contradiction label sequence; constructing an evolution state vector, inputting the evolution state vector into an improved GMN model, and generating a structure repair state vector through a structure guiding unit, a deviation sensing unit, a path fusion unit and a node evolution unit in sequence; and executing result updating and conflict rearrangement, completing standard template alignment and semantic coverage verification, and generating a final structure recognition result. According to the method, the structural integrity, the semantic consistency and the automatic repair capability of directory recognition are improved, and the method has relatively high adaptability and intelligent level.

The invention discloses an aeronautical ice accretion identification method based on multi-source data, which solves the coverage blind area of a single data source (mode / satellite) through multi-source data fusion processing, improves the integrity of a three-dimensional space, and further establishes a scene classification foundation through cloud layer structure identification and accurate division of a cloud area / a precipitation area. Through membership function construction processing, a nonlinear relation between meteorological elements and icing is quantified, and a traditional threshold value method is replaced; furthermore, through physical scene classification processing, single-layer cloud / multi-layer cloud / freezing water freezing mechanisms are distinguished, and misjudgment is avoided; through initial index calculation, CTTmap is dynamically adjusted according to different scenes, and satellite detection deviation is corrected; finally, through dynamic index correction processing, vertical motion, supercooled water and a real-time report are introduced, and result self-adaptive optimization is achieved.

The invention discloses a sewage high-risk pollutant screening and identification method based on fragmentation tree pre-training, and the method comprises the steps: constructing a fragmentation tree pre-training data set according to the second-level mass spectrum data of known high-risk pollutants, so as to carry out the self-supervision pre-training of a graph neural network encoder, and obtaining a fragmentation tree editor; the method comprises the following steps: acquiring secondary mass spectrum data of suspected high-risk pollutant related compounds in a to-be-detected sewage sample, constructing a fragmentation tree set, constructing a high-risk pollutant screening model based on a fragmentation tree encoder, performing high-risk pollutant signal screening on the fragmentation tree set to obtain a high-risk candidate fragmentation tree set, and performing high-risk pollutant signal screening on the high-risk candidate fragmentation tree set. And generating a candidate molecule set of each high-risk candidate fragmentation tree, and identifying a target molecular structure and a matching score of the high-risk candidate fragmentation tree from the candidate molecule set. According to the invention, rapid screening and priority ranking of high-risk pollutants in sewage are realized, and candidate output of structure identification is further provided on the basis of screening.

The invention relates to the technical field of government affair intelligence, in particular to a government affair application platform based on an artificial intelligence model, and the platform comprises a semantic recognition module, a tag affiliation module, a rule screening module, a path generation module and a path verification module. According to the method, the subject-called structure for recognizing noun phrases and approval verbs in the government affair approval text is adopted, semantic boundaries are defined, the structure recognition accuracy is enhanced, keyword and field item comparison adopts a field-level consistency mode, the tag affiliation precision is improved, field value comparison logic is introduced in rule node screening, and the accuracy of tag affiliation is improved. In path generation, a path structure with clear logic is established through sorting index and field dependence identification, field conflicts and path chaos are avoided, field value consistency verification ensures that path content is highly matched with an approval text, and it is guaranteed that the path is effective and available; automatic processing from text analysis and field affiliation to path verification is achieved, and the efficiency of the government affair approval process is improved.

The invention provides an interaction optimization method and system based on a big data language model, and is applied to the field of interaction data processing. According to the method, by introducing grammar structure recognition, task structure diagram construction and interaction intention chain generation mechanisms, the technical defect that a big data language model lacks structured recognition and semantic decomposition capabilities when processing a compound request is overcome, on the basis of judging the model capability, grammar features in an input request can be recognized, and the interaction intention chain generation efficiency is improved. The method comprises the following steps of: establishing a clear task node and a dependency path, further extracting a structured intention chain comprising elements such as actions, objects, conditions, purposes and the like, generating a matched execution structure, and finally establishing a parallel and nested intention request structure by virtue of data dependency between an interactive intention trigger word and a semantic instruction. It is ensured that the language model can stably process multi-stage natural language requests with complex logic, and the overall interaction quality and the execution effect are improved.

The invention discloses a safety detection system for a high-speed rail transit corollary equipment structure. The safety detection system comprises an excitation response module, an infrared heat source image acquisition module, a parameter monitoring module, an image preprocessing module, a heat source feature extraction module and a structure anomaly recognition module. In the train operation process, the system excites trackside equipment structure response through train body aerodynamic disturbance, thermal response images are collected through high-frame-rate infrared imaging, and image standardization processing is carried out based on disturbance intensity and the relative distance. And the system further extracts thermal response characteristics of the equipment structure area, and compares the thermal response characteristics with a normal template to realize abnormal structure identification and positioning. Under the condition that train operation is not affected, non-contact, high-precision and intelligent detection of the structure state of the trackside equipment is achieved, and the method has the advantages of being flexible in deployment, accurate in recognition, high in real-time performance and the like and is suitable for rail transit operation and maintenance scenes.

The invention relates to the technical field of intelligent medical treatment, and discloses a cataract surgery navigation control method based on multi-modal fusion, which comprises the following steps: carrying out lightweight preprocessing on an eye fundus image, ultrasonic probe data, a corneal topographic map and intraocular pressure parameters through an edge computing architecture, and constructing an efficient feature extraction network by adopting a knowledge distillation technology; self-adaptive feature fusion is realized based on an uncertainty quantification mechanism, real-time structure recognition is provided by utilizing an augmented reality technology, an intelligent navigation strategy is generated by pre-training a reinforcement learning model, an instrument control system with multilayer safety guarantee is established, and continuous optimization of the system is realized by adopting an online learning mechanism. According to the invention, the precision and safety of the operation are effectively improved, and an effective technical scheme is provided for intelligent navigation control of the cataract operation.

The invention relates to an AI intelligent document processing method based on a large language model, and the method comprises the steps: receiving the input of an original document, and obtaining the text features, visual layout features and semantic structure features of the document through a multi-modal feature extraction module; processing the multi-modal features by using a layout-aware attention mechanism, and generating an incidence matrix of document elements; inputting the extracted features into a pre-trained large language model, and generating an intermediate representation containing an association relationship between document elements; and dynamically selecting a processing path according to task requirements, wherein the processing path comprises but is not limited to document classification, information extraction, content generation or question and answer response. According to the AI intelligent document processing method based on the large language model, through an attention mechanism of layout perception and an experiment on a standard data set, table structure recognition F1-score is improved, the cross-page element association accuracy is greatly improved, the vision-text feature alignment loss is reduced, the processing time of a 100-page PDF document is greatly shortened through a dynamic partitioning processing strategy, and the processing efficiency of the 100-page PDF document is greatly improved. And the memory occupancy peak value of the GPU is reduced.

The invention relates to the technical field of medical image processing, in particular to a tumor image classification method and system based on deep learning, and the method comprises the following steps: obtaining a tumor image gray-scale map, positioning an edge, recognizing a closed structure region, dividing tumor region image blocks, calling a standard image block, analyzing the contour and texture offset degree, and carrying out the layering comparison of labels. And reversely marking image blocks, analyzing edges and consistency of a detail layer, a texture layer and a structure layer, constructing a collaborative matrix to carry out multi-task mapping, and outputting a tumor classification set. According to the method, a closed structure area is positioned and extracted through gray boundary change, the initial recognition precision is enhanced, non-structural interference is reduced, a standard image block is mapped to realize morphological comparison, the classification distinguishing force is improved, label collection prediction errors are compared in a layered manner, error saliency image blocks are dynamically labeled, and a more sensitive feedback mechanism is realized; the multi-layer division enhances the hierarchical expression of the features, enhances the multi-scale aggregation capability, and improves the classification accuracy and the structure recognition generalization capability.

The invention relates to an FPGA-based GPR ground penetrating radar electromagnetic signal synchronous high-speed data acquisition system and method. The system comprises a front-end signal processing module, a high-speed ADC sampling module, a clock synchronization control module, an FPGA real-time processing module, a target detection and feature extraction module, a mechanism data storage module and a remote data transmission module. On the engineering safety level, the service life of equipment in a strong electromagnetic environment is prolonged by three times through the multi-stage protection design, and the annual failure rate is reduced to 0.1% or below; an ultra-high-speed sampling system (83.2 GS / s) is combined with femtosecond-level clock synchronization, so that 0.5 mm-level thin-layer structure recognition becomes possible, and a technical foundation is provided for geological disaster early warning; a pioneering photoelectric double-channel transmission mechanism is adopted, the electromagnetic interference suppression is improved by 12dB while the 1ps synchronization precision is kept, and the traditional contradiction between high-speed sampling and signal fidelity is broken.

The invention relates to the technical field of computer vision and document analysis, and discloses a table structure identification method and system based on a self-adaptive anchor frame, a terminal and a medium. The method comprises the following steps: acquiring multiple types of typical table image samples, labeling structured data and merging types of table cells, and constructing a travel channel data set and a column channel data set through multi-modal processing; constructing a table structure recognition model, wherein the model comprises a trunk feature extraction network, a double-branch detection head and a merging cell classification module; training a table structure recognition model based on the row channel data set and the column channel data set; and executing a structure identification task on an input table image by utilizing the trained model, and generating structured data containing physical coordinates and logic position relationships of the cells according to an identification result. According to the method, the problem of accumulative errors caused by asymmetry of row and column detection logic in a traditional method can be effectively solved, and meanwhile the problem that the row width and the column height are inconsistent due to complete dynamic adjustment is avoided.

The invention discloses a mining area disaster control-oriented dual-stage scheme construction method, and relates to the technical field of mine safety engineering, and the method comprises the following steps: S100, constructing a high-resolution multi-mode joint detection system, and carrying out the cooperative arrangement of a seismic wave array and an electrical method profile, and synchronously collecting the wave impedance response and electrical property anomaly characteristics, and obtaining a physical anomaly information set covering full depth. According to the method, through multi-mode joint detection and heterogenous data fusion, the precision and integrity of deep structure recognition are improved, and in combination with geological evolution reasoning and dynamic inversion, recognition blind areas are complemented, and accurate positioning of high-risk disaster-causing units is realized. Based on matching of a disaster index model and a parameterization strategy, automatic mapping of disaster partition and a treatment scheme is achieved for the first time, scientificity, targeting and resource allocation efficiency of disaster treatment are remarkably improved, and the method has important engineering application and popularization value.

The embodiment of the invention discloses a message transmission method and device, and the method comprises the following steps: carrying out the structure recognition and semantic segmentation of original message data based on a pre-configured partitioning strategy, and generating a plurality of pieces of partitioning data with the structural integrity; generating a metadata record containing a task identifier, a block index, a total block number, a structural hash abstract and a receiving timestamp for each piece of block data; performing data normalization processing on each block of data, wherein the processing comprises the steps of unifying character codes and removing line feed characters and space characters; and sending the data of each block in a transaction mode through a Kafka message queue, and dynamically adjusting the block retry sending time interval and the upper limit of the number of times according to packet loss and delay indexes. According to the embodiment of the invention, the business service can be effectively decoupled, the throughput of the system is remarkably improved, efficient sorting and complete recombination of the block data are realized, and the transmission reliability is further improved.

The invention belongs to the technical field of character pattern recognition, and particularly provides an oracle structure recognition method of a font pattern isomorphic network, which comprises three stages of font pattern structure feature extraction, font pattern isomorphic network model and oracle font structure recognition. Through oracle font skeleton extraction, oracle font skeleton singular point detection, skeleton burr and bifurcation optimization, and key point extraction as a node communication path as an edge, an oracle font graph structure is established, and graph structure nodes, edges and graph level features are calculated. The font graph isomorphic network integrates the advantages of a graph isomorphic network and dynamic edge and graph embedding, and learns vector representation of multi-level features such as nodes, edges and graphs of the oracle font graph structure so as to train and identify the font structure of the oracle. The method can be used for structure matching recognition of oracle character pattern images, has the advantage of being high in recognition accuracy, and is suitable for recognition of character pattern structures such as gold texts and seal scripts and recognition of social networks, biological information and molecular structures.

The invention provides a battery pack active equalization control method, device and equipment and a storage medium, and the method comprises the steps: carrying out the structure recognition of single cells in a battery pack according to the cell configuration information, and generating a cell mapping table; according to the state parameter of each single battery cell in the battery cell mapping table, determining the residual electric quantity value of the corresponding single battery cell; performing pairing calculation on the residual electric quantity values, determining a packet-crossing residual electric quantity difference value between the target battery cell and the mapping battery cell, and generating a corresponding scheduling candidate set according to the packet-crossing residual electric quantity difference value; constructing a cross-packet equalization path set according to the scheduling candidate set, and determining a target cross-packet equalization path meeting all equalization triggering conditions; and performing electric quantity balance control between the main battery pack and the auxiliary battery pack according to the target pack-crossing balance path. According to the scheme, the electric quantity between the main battery pack and the auxiliary battery pack can be balanced through the pack-crossing balancing path under the multi-battery-pack structure, the problem of over-charging or over-discharging of the batteries is avoided, and the service life of the batteries is guaranteed.

The invention relates to the technical field of medical image processing, in particular to an ultrasonic image segmentation method based on edge guidance, and the method comprises the steps: inputting an ultrasonic image into an edge extraction branch and an image encoder; outputting an edge mask image corresponding to the image through the edge extraction branch; performing noise suppression and topological correction on the edge mask image to generate a closed edge constrained by an anatomical structure; generating prompt box information based on the closed edge, and inputting the prompt box information to a prompt encoder; fusing the image prompt features output by the prompt encoder, the image features extracted by the image encoder and the edge mask information; the fusion features are input into a decoding module, a final segmentation result is obtained, the synergistic effect of edge information and visual prompt is fully utilized, the perception ability of the model for the anatomical structure in the ultrasonic image is improved, and the accuracy and robustness of segmentation are remarkably improved. Therefore, the problems of fuzzy boundary, inaccurate prompt, weak structure identification capability and the like in related technologies are solved.

The invention relates to the cross technical field of computer vision and natural language processing, and particularly provides a zero sample template inference and document structured recognition method and device, and the method comprises the following steps: S1, document image collection and preprocessing; s2, performing layout sensing partitioning and position coding; s3, priori or example information is constructed and injected; s4, performing cross-modal fusion and expression construction; s5, performing automatic format analysis and field slot filling; s6, generating a cue word-free extraction instruction; s7, performing field area parallel character recognition; s8, performing semantic verification and result standardization; and S9, outputting the structured field-value data. Compared with the prior art, the method has the advantages that the dependence of a traditional method on template making, cue word writing and large-scale sample training can be avoided, the flexibility, accuracy and online speed of document structured recognition are remarkably improved, and the method has good intelligence and rapid adaptation capacity and is suitable for diversified document recognition scenes.

The invention relates to the technical field of document intelligent review, in particular to a multi-mode digital publishing intelligent review system and method based on a large model, and the system comprises a sample construction module, a multi-mode analysis module, a structure recognition module, a term verification module and an annotation output module. According to the method, a model configuration data set is generated by analyzing a subject-object combination relationship in a text and jointly screening part-of-speech, sentence pattern and semantic structure, the accuracy of semantic recognition and structure judgment is improved, and local features of semantic offset and image-text expression separation are recognized by combining a comparison relationship between an image action path and a text behavior object; a paragraph logic structure and a theme connection mode are analyzed based on semantic offset information, the problems of content dislocation and theme jump between paragraphs are effectively revealed, matching change tracks and part-of-speech continuation fluctuation in term cross-paragraph contexts are recognized, label overlapping and coverage redundancy conditions are analyzed from sentence group hierarchy, and the term cross-paragraph contexts are obtained. And forming a label combination suggestion and constructing an annotation structure record.

The invention discloses a health early warning management system and method based on physical examination information, and particularly relates to the field of skeleton structure recognition, and the method comprises the steps: constructing a structural posture imaging data set based on a physical examination posture image of a student, extracting three-dimensional structure coordinates of parts, such as a scapula, a spine and a pelvis; and uniformly mapping to a skeleton reference system to calculate a body inclination index, a spine offset rate and a pelvic inclination index. Then, constructing a time sequence of posture structure parameters, and generating a trend feature vector sequence according to academic difference and gradient extraction; and further constructing a posture structure evolution trajectory diagram by taking the structure parameters as nodes and the trend vectors as edge attributes, and capturing a cross-year posture change path of the students. On the basis, all paths are aligned according to ages, the sudden change trend is extracted, the posture deviation high-incidence risk of a specific age group is recognized, and accurate student posture early warning management is achieved.