168 results about "Conditional random field" patented technology

The invention provides a tunnel crack identification method and system based on multi-source image processing, and relates to the technical field of tunnel engineering, and the method comprises the steps: obtaining multi-source data of a horizontal rock stratum tunnel; performing spatial registration on the multi-source data to generate a multi-modal image under the same reference system; extracting multi-modal features according to the multi-modal image, and constructing a multi-source feature map under the same space grid; performing crack initial detection on the multi-source feature map in combination with multi-scale filtering and structure tensor analysis to obtain a crack candidate region mask; accurate crack identification is carried out through the crack candidate area mask, a crack identification result is obtained by combining bedding direction constraint and a microconditional random field, and the crack identification result comprises a crack segmentation map and a crack category. The method solves the problem that existing crack identification does not consider the bedding characteristics of the horizontal rock stratum tunnel.

The invention discloses a mobile terminal streetscape image real-time segmentation method based on a lightweight neural network, and relates to the technical field of image segmentation. The method comprises the following steps: firstly, carrying out 320 * 320 adjustment, Z-score standardization, adaptive histogram equalization and 3 * 3 Gaussian filtering preprocessing on an input streetscape image; then, an improved MobileNetV3 backbone network is used, and a five-scale feature map is output in combination with DropBlock regularization through eight feature extraction stages including depth separable convolution and an SE attention module; multi-scale features are fused through a U-shaped structure, and a fusion feature map is generated through up-sampling, element-by-element addition of dimension reduction low-layer features and an attention gating module; and during reasoning, outputting a segmentation mask by using a convolutional layer, Softmax and a conditional random field, and finally performing knowledge distillation, weight pruning, 8-bit quantization and TensorRT optimization. According to the invention, high-precision real-time street view segmentation is realized, the robustness is high, and the method is suitable for different devices and scenes.

The invention relates to the technical field of remote sensing image processing, and discloses a remote sensing image adaptive identification method and system for territorial space planning, and the method comprises the steps: obtaining a multi-source remote sensing image data set of a research region, feature extraction, cloud detection, quality evaluation and adaptive preprocessing are carried out; carrying out prototype network coding, calculating a category prototype and probability, and supporting fine tuning of a set; carrying out multi-scale cavity convolution and category scale attention fusion; evaluating the adaptability score of the comprehensive fusion feature map set, and carrying out weighted fusion, classification and normalization; change detection is carried out, stable and change regions are segmented, and time sequence context features are extracted and constrained optimization is carried out; entropy is fused, a boundary is decided, uncertainty is estimated, and weighted fusion is carried out according to a change area; conditional random field optimization, confidence level grading and connected domain identification are carried out; the automation level, the adaptive capacity and the recognition reliability of remote sensing monitoring of territorial space planning are improved.

The invention discloses a field large model geological survey report generation method based on a knowledge graph, and the method comprises the following steps: S1, building an engineering survey field data set through multi-source heterogeneous data collection and structured preprocessing, and the engineering survey field data set comprises five text dimension tags divided according to engineering survey specifications; s2, constructing an engineering investigation knowledge graph; and S2A, knowledge extraction, wherein a bidirectional encoder presentation layer-bidirectional long short-term memory network-conditional random field joint extraction model is adopted. According to the method, a bidirectional encoder presentation layer-bidirectional long short-term memory network-conditional random field joint extraction model is improved on knowledge modeling to perform high-precision entity-relation joint extraction, a geological knowledge map with consistent semantics and clear structure is constructed based on RDF, rule reasoning and graph neural network reasoning mechanisms are fused, and the method has the advantages of high-precision entity-relation joint extraction and high-precision entity-relation joint extraction. Deep mining and complementation of explicit and implicit knowledge are realized, and the ability of the prior art in knowledge expression granularity and reasoning breadth is improved.

The invention provides a method and equipment for constructing a text entity recognition model based on large model data enhancement. The method comprises the following steps: firstly, constructing an initial model comprising a preprocessing unit, a syntax dependency analysis unit, a data correction enhancement unit, a context coding unit, a syntax enhancement unit, an expression fusion unit and a sequence decoding processing unit; preprocessing the training sample to obtain a preprocessed text, and establishing a preliminary dependency graph through syntactic analysis; correcting and enhancing the text and the dependency graph by the large language model to obtain a text sequence and a dependency graph for subsequent use; encoding the text sequence to obtain an initial representation vector, and enhancing the vector in combination with the dependency graph; after fusion, decoding and outputting a prediction label containing a lexical entity label; and calculating loss by using a function containing conditional random field structure loss, judging convergence, and if not, updating parameters and continuing training until a target model is obtained. According to the method, data are optimized and expanded by means of a large language model, syntactic dependency enhancement features are combined, and the recognition capability of the complex text named entities is improved.

The invention relates to the technical field of intelligent breeding, and discloses a beef cattle identification method based on multi-scale segmentation optimization and multi-modal data fusion, which comprises the following steps: acquiring beef cattle multi-modal image data, carrying out time-space synchronization and fusion to construct a fusion image, carrying out preprocessing of illumination invariant transformation and multi-scale pyramid construction, and carrying out multi-scale segmentation optimization and multi-modal data fusion on the fusion image; generating an enhanced image, inputting the enhanced image into a depth-guided attention segmentation network to extract double-branch foreground features, generating an initial segmentation probability graph, optimizing a segmentation mask in combination with a conditional random field model and a motion consistency constraint, finally extracting multi-dimensional features from the mask, inputting the multi-dimensional features into a multi-classification support vector machine to perform individual classification reasoning, and obtaining a final segmentation result. And obtaining a beef cattle individual identification result. Therefore, the foreground and background segmentation method applied to the intelligent breeding scene for beef cattle individual recognition is provided, the beef cattle individual recognition precision is improved, and meanwhile the management requirements of commercial breeding for high precision, high stability and multi-scene adaptability are met.

The invention relates to the cross technical field of artificial intelligence and medical image processing, in particular to a medical image segmentation system and method based on a wavelet bridge diffusion model and an efficient conditional random field. A WBDM-ECRF framework is constructed and comprises a discrete wavelet transform module, a BDM-T module, a BDM-S module and an ECRF module; decomposing the image through discrete wavelet transform, extracting a low-frequency sub-band, and enhancing the contrast ratio of a focus and normal tissues; the BDM-T takes U-Net as a backbone, integrates a FlashAttention mechanism, and optimizes a variance formula to realize efficient training; the BDM-S adopts a leapfrog sampling strategy, so that the reasoning time is greatly shortened; the ECRF introduces a multivariate potential function of a structural similarity index and smooth operation through edge expansion, and accurately optimizes edge segmentation. According to the method, the dependence of marked data is reduced, the training and reasoning efficiency is improved, the problem of fuzzy edge segmentation is solved, the Dice coefficient and intersection-union ratio performance on the ISIC data set is excellent, and reliable quantitative support is provided for disease diagnosis and treatment.

The invention relates to a seismic liquefaction assessment method based on conditional random field simulation, which comprises the following steps: firstly, obtaining a logarithmic normal distribution random field of a target area under a corresponding SPT-N value, then resampling through a Bootstrap method, constructing a weighted prior probability density function of the target area in combination with a likelihood function, and finally calculating the seismic liquefaction of the target area according to a Bayesian theory. A Markov chain Monte Carlo sampling method is combined, through posterior probability density distribution, an optimal horizontal direction correlation distance is determined, a covariance matrix is constructed to generate a conditional random field, and then through multiple times of simulation, the conditional random field is converged; and finally, aiming at the target area, through calculation of a cyclic stress ratio and a cyclic resistance ratio, constructing a liquefaction probability distribution diagram corresponding to the target area. According to the method, a conditional random field simulation method is inferred and improved by combining Bootstrap and Bayesian theories, the precision and reliability of geological parameter simulation are remarkably improved, and reliable data support is provided for seismic liquefaction assessment of deep and uneven site engineering.

The invention provides an image video retrieval method based on a domain fine-tuning large language model, which comprises the following steps: performing fine-tuning on a pre-training model to obtain a fine-tuning pre-training model for intention classification and keyword extraction; performing dynamic iteration screening on an optimal prompt template through Monte Carlo tree search in combination with a hidden Markov model (HMM); performing noise filtering on the keyword list, and predicting category labels of the filtered keywords through a conditional random field model to obtain a keyword enhancement set; combining with the user intention to generate a query condition, and obtaining a candidate resource set; and according to the similarity between the user query text and the candidate resource set, and in combination with the optimal prompt template, obtaining the resource path with the highest matching score between the user query and the candidate resource, and obtaining the retrieved image or video, so that the identification deviation possibly occurring when a general model processes proper nouns and terminologies can be effectively solved, and the user experience is improved. And the retrieval accuracy and response speed are improved, so that the retrieval accuracy and professional adaptability are improved.

The invention relates to a colorectal cancer MRI image segmentation method and system based on multi-dimensional feature fusion, and relates to the technical field of medical image processing. The method comprises the following steps: firstly, collecting multi-source data of a colorectal cancer patient, preprocessing the data, and constructing a multi-scale feature extraction module to obtain feature data of an image; through a multi-dimensional feature fusion module, features of seven dimensions including texture, shape, gray scale, modal association, life habits, functional metabolism and tissue specificity are integrated, and feature screening and weight distribution are realized in combination with a space-channel attention mechanism; and finally, carrying out feature decoding through a U-Net network, and optimizing a segmentation boundary by adopting a full-connection conditional random field (CRF) of an adaptive potential function. According to the method, multi-modal data integration, multi-scale feature extraction, multi-dimensional feature fusion, attention mechanism enabling, U-Net network decoding and CRF post-processing are combined, so that the accuracy and robustness of colorectal cancer MRI image segmentation are improved.

The invention relates to the technical field of computer vision and image processing, in particular to a cross-day-and-night boundary target thermo-optic joint detection method, which comprises the following steps of: firstly, acquiring a visible light image and an infrared image and unifying the visible light image and the infrared image to an image plane reference system; extracting topological features and multi-scale energy features from the input, fusing the topological features and the multi-scale energy features to generate a feature map, and outputting a central heat map, target size regression and sub-pixel offset by adopting anchor-frame-free detection; implementing optimal transmission correction according to an evidence field obtained by normalization of a cross reconstruction residual field to obtain a correction heat map and an initial candidate; triggering a fixation area by using the shape correction heat map and the initial candidate, executing super-resolution and secondary detection in the area, and performing affine reprojection and primary detection fusion to form an updated heat map and a fusion candidate; and in combination with uncertainty and topological consistency, a final detection set is output by using a conditional random field and non-maximum suppression. The method is stable in low-contrast, small-target and strong-interference scenes.

The invention discloses an agent intention understanding method based on multi-modal information fusion, relates to the technical field of natural language processing, and solves the technical problem that multi-agent intention understanding cannot be realized in a dynamic environment. Voice, text and visual features are mapped to the same hidden space through maximum mean value difference minimization, distribution difference between modes is eliminated, joint features in the hidden space can capture cross-mode complementary information at the same time, and the richness and anti-noise capability of feature expression are enhanced. CRF (Conditional Random Field) is combined with a special dictionary in the scheduling field to carry out candidate set expansion on fuzzy keywords, and specific semantic variants in the field are captured. The optimal grammar of the scheduling instruction is analyzed and decomposed into a feature vector causal relationship pair, and components of an intention and logic dependence of the intention are clarified. A dynamic knowledge graph is constructed based on the causal relationship, probability dependence among parameters is quantized by using a Bayesian network, the causal relationship strength is dynamically adjusted, and environmental changes are adapted.

The invention relates to the technical field of natural language processing, in particular to a foreign trade letter intelligent semantic error correction and business term optimization method and device, and the method comprises the steps: obtaining a target foreign trade letter, and constructing a target corpus; performing Chinese word segmentation and part-of-speech tagging on the target foreign trade letter; performing term optimization based on the word segmentation result and the knowledge graph; identifying the letter title by using a conditional random field model, and converting the letter title into structured data; a Bi-LSTM-CRF model is adopted to carry out risk point detection, including Bi-LSTM coding, feature engineering and Max-pooling technologies, a part-of-speech sequence is obtained through a softmax function and a Viterbi path, and sequence labeling is carried out to obtain a risk point detection result; and finally, performing Chinese error correction based on the word segmentation result after part-of-speech tagging and the knowledge graph. The recognition and correction accuracy of foreign trade terminologies is improved, and communication obstacles caused by nonstandard use of the terminologies are effectively reduced.

The invention discloses a rock slope multi-mode instability probability analysis and risk evaluation method based on a drilling constraint condition random field, and belongs to the technical field of geotechnical engineering and engineering geology. According to the method, the conditional random field which is strictly consistent with the actually measured data is constructed under the condition of limited drilling exploration data, so that the reasonable expression of the spatial heterogeneity of the rock mass mechanical parameters is realized, and the authenticity of a numerical model and the reliability of an analysis result are greatly improved; the defect of insufficient result credibility caused by lack of actual measurement constraint of the unconditional random field is effectively overcome; through a large amount of multi-sample numerical simulation, multiple instability modes possibly occurring on the rock slope under different parameter space distribution conditions can be identified, the occurrence probability of each instability mode is quantitatively calculated, various risks possibly existing on the slope can be comprehensively grasped, only the most dangerous single condition is concerned, and the risk assessment accuracy is improved. And the accuracy and engineering applicability of slope stability analysis and risk evaluation are improved.

The invention discloses an electric power project material inventory intelligent generation method and system based on multi-modal document analysis and knowledge graph semantic mapping, and belongs to the field of artificial intelligence and electric power engineering. Comprising the steps that in the first stage, unified analysis and semantic coding are conducted on text documents, engineering drawings and table data, and a semi-structured original information set is generated; in the second stage, named entity recognition and type labeling are carried out on the original information set based on a two-way long-short term memory network and a conditional random field hybrid model, a structured entity set is generated, an electric power material field knowledge graph is constructed on the basis, and a standard key point rule base is established; semantic mapping and information completion are carried out through a multi-level matching mechanism and knowledge reasoning, and a standardized material inventory is generated after quality control. According to the invention, intelligent and automatic generation of the electric power project material inventory is realized, and the accuracy and efficiency of material inventory generation are improved.

The invention belongs to the technical field of image recognition, and particularly relates to an automatic recognition and processing method for weed shielding between photovoltaic arrays. The method comprises the following steps: step 1, data acquisition, polarization highlight removal and candidate region generation; step 2, calculating an expected shadow direction and an expected shadow length according to the sun azimuth, and carrying out direction consistency judgment; step 3, performing Robust PCA sparse extraction on the mismatch set and performing form guidance enhancement on the direction consistent set, and merging processing results to generate a weed foreground set; and 4, constructing a conditional random field graph model on the component grid, refining the boundary, and outputting a weed mask and severity. According to the method, weeds and shadows can be stably distinguished in the environments of strong illumination, high reflection and shadow staggering, the recognition precision and robustness are effectively improved, the conversion from pixel-level segmentation to component-level risk quantification is realized, the manual inspection cost is remarkably reduced, and the operation and maintenance efficiency and the power generation stability of a photovoltaic power station are improved.

The invention discloses a colorectal gland segmentation method and device, equipment and a storage medium, and the method achieves precise segmentation through three-stage weak supervised learning: in the first stage, constructing a self-supervised data set by using unlabeled colorectal pathological image blocks, and carrying out the pre-training and fine tuning of a visual model, and obtaining a feature encoder adaptive to the pathological tissue field; in the second stage, an original data set is constructed based on a pathological image with an image-level gland tag, and a high-quality fine pseudo tag is generated through feature extraction, classification network training, attention heat map generation, image fusion thresholding and full-connection conditional random field optimization; and in the third stage, a weak supervision segmentation data set is constructed by using a fine pseudo label, a boundary perception loss function is adopted to train a gland segmentation model, and finally pixel-level gland segmentation of the unlabeled colorectal pathology image is realized. According to the method, accurate gland segmentation can be realized without pixel-level labeling, and the labeling cost is effectively reduced.

The invention provides a multi-mode command entity recognition method based on course comparative learning, and belongs to the technical field of natural language processing and computer vision. The method comprises the following steps: respectively carrying out semantic representation modeling on text and image data to obtain text features and image features; multi-level semantics from coarse granularity to fine granularity between the text features and the image features are aligned through staged comparative learning, and association between the text features and the image features is enhanced; fusing the text features and the image features through a gating multi-interest fusion mechanism to obtain fusion features; and performing sequence labeling on the fusion features by using a dynamically weighted conditional random field model. By combining a learnable gated multi-interest fusion module and a category-sensitive dynamic decoding mechanism, the problems of modal noise interference and category imbalance are effectively relieved, and the accuracy and robustness of multi-modal entity recognition and the applicability and stability of the model in a real social media environment are improved.

The invention belongs to the technical field of artificial intelligence and medical information processing, and particularly relates to a medical text big data mining system based on natural language processing, which comprises a medical field pre-training module, a text structured extraction engine module, a knowledge graph guide reasoning module, a clinical path optimization module and a medical question and answer and literature mining platform module, according to the system, a self-adaptive hierarchical attention mechanism is adopted, and medical texts of any length are processed through dynamic fusion of local attention and global attention; constructing an entity dependency relationship graph and coding semantic association between entities by adopting a graph neural network; introducing a medical ontology knowledge graph to carry out conditional random field joint reasoning, and correcting unreasonable annotations; the optimal diagnosis and treatment mode is mined and recognized through the frequent mode, the entity recognition F1 value reaches 96.2%, the long-distance relation recognition accuracy reaches 91%, the structured data accuracy reaches 93.7%, the data preparation time is shortened by 85%, and the document review efficiency is improved by 320%.

The invention discloses a DBSCAN unmanned aerial vehicle path planning method based on depth estimation and optical flow analysis, and relates to the technical field of unmanned aerial vehicle autonomous navigation, and the method comprises the steps: extracting an image depth feature through a VGG16 model, carrying out the enhancement and segmentation of an image through Gamma transformation and a region growing algorithm, optimizing a seed group through optical flow analysis and a conditional random field algorithm, and carrying out the optimization of the depth feature of the image. A DBSCAN algorithm is used for carrying out path clustering, an A * algorithm is used for carrying out obstacle avoidance path planning, and a double-loop PID control algorithm is combined for execution. According to the invention, through fusion of multi-modal data and path clustering optimization, navigation precision and robustness in a complex environment are improved, seed group dynamic updating is realized based on seed probability calculation and optical flow analysis of a conditional random field, sensing precision, environmental adaptability and real-time reaction capability are improved, and through combination of double-loop PID control and an A * algorithm, the real-time response capability of the system is improved. And the real-time obstacle avoidance capability is optimized.

The application provides a machine learning-based financial account allocation and anomaly detection method and system, relates to the technical field of machine learning, and comprises the following steps: a causal diagram model is constructed, a conditional random field is used to generate a time sequence attention weight matrix to perform weighted fusion, and a difference reason is attributed; a reinforcement learning method is used to process a difference feature vector, a processing strategy is generated and optimized, and finally a difference processing scheme is output. The application can improve the account accuracy, automatically find abnormal transactions, reduce the manual processing cost, and effectively prevent financial risks.

A lightweight network-based driving behavior analysis method based on multi-source feature extraction includes: First, predicting the lane-changing frequency of vehicles using conditional random fields as features representing external influences of physical constraints and lane interference. Second, extracting multiple features from vehicle driving information, vehicle physical characteristics, lane location and probability, and traffic conditions to obtain the weight of each feature, thereby improving the accuracy of driving behavior analysis and rating. Finally, simultaneously utilizing a lightweight network for feature fusion to achieve the purpose of driving behavior analysis and rating. Because of the use of a lightweight network, the model training time is shortened, enabling real-time response in behavior analysis. This method features convenient model training, strong operability, and resource saving.

The invention provides a parallelization boundary perception cloud or shadow rapid segmentation method, and belongs to the technical field of image processing, and the method comprises the steps: giving a remote sensing image global observation value and a preset unary potential function of a cloud / shadow coarse mark, and carrying out the iterative segmentation through employing a parallelization boundary perception segmentation method; the initialization of a Gaussian KD tree is carried out; iteratively using the following steps to carry out conditional random field parallelization rapid message passing, class compatibility transformation of non-normalized probability, local updating of non-normalized probability and probability normalization; and after iterative convergence, optimal classification is obtained. According to the method, the sensitivity to the low-level visual features of the region of interest in the image is enhanced, the segmentation model is guided to perform accurate segmentation on the edge of the target of interest, and the target edge is fully compact and sharp in segmentation prediction; and designing a Gaussian KD tree-based parallelization model implementation method to solve the problem of rapid calculation of a boundary perception segmentation model.

This invention relates to the field of named entity recognition technology, and discloses a multimodal named entity recognition method and apparatus based on deep interaction of text and image features. The method includes: acquiring original text and an original image; extracting multimodal features from the original text and the original image respectively to obtain a text feature sequence and an image feature sequence fused with multi-scale features; performing cross-modal deep interaction fusion on the text feature sequence and the image feature sequence to obtain an interacted multimodal feature sequence; performing context-aware adaptive gating fusion on the interacted multimodal feature sequence to obtain a fused feature sequence; and performing conditional random field sequence decoding on the fused feature sequence to obtain an entity label sequence. This invention achieves deep bidirectional interaction and dynamic adaptive fusion of multimodal information, effectively solving the text ambiguity problem and improving the accuracy of named entity recognition.

This invention discloses an environmental anomaly broadcasting method and system based on semantic analysis and knowledge graph, belonging to the field of environmental anomaly detection technology. Key technical points include: acquiring the current inspection record text and performing semantic analysis to obtain text keywords; determining the nodes corresponding to the text keywords in a preset knowledge graph based on a conditional random field (CRF) model; training the CRF model according to a preset constraint matrix; the types of nodes in the knowledge graph include enterprises, emission outlets, processes, pollutants, and sensitive targets; obtaining anomaly information corresponding to the current inspection record text based on the association relationship between the nodes corresponding to the text keywords and nodes of each type in the knowledge graph; and broadcasting the anomaly information. This invention improves the accuracy of node positioning and generates more accurate anomaly information by constructing a constraint matrix to clarify the matching rules between text and nodes and training the CRF model accordingly.

The application provides a metadata intelligent identification method and device based on semi-supervised learning, and the method comprises the following steps: generating a metadata keyword identifier according to a conditional random field, wherein the metadata keyword identifier is used for extracting metadata features corresponding to any identification; acquiring an initial labeled data set, training the metadata keyword identifier according to the initial labeled data set to generate a metadata classifier; acquiring an unlabeled data set, predicting the unlabeled data set according to the metadata classifier to generate a prediction result; generating an intermediate training data set according to the prediction result; cyclically self-training the metadata keyword identifier and the metadata classifier according to the intermediate training data set to obtain a cyclically self-trained metadata classifier; and identifying metadata intelligently according to the cyclically self-trained metadata classifier. Through the application, the problem of low metadata identification construction efficiency in the related art is solved.

Embodiments of the present application provide a sensitive entity recognition model training method, a sensitive entity recognition method, a computing device, a computer readable storage medium and a computer program product. The sensitive entity recognition model training method comprises: performing word segmentation processing on unstructured text and generating a sample label sequence based on a preset rule annotation; inputting a word segmentation sequence into a bidirectional encoding layer to generate a semantic vector, so as to determine an emission score of each label corresponding to the word segmentation; inputting the emission score into a conditional random field layer to obtain a transition score between adjacent labels, and combining the emission score and the transition score to determine a joint probability of each label corresponding to the word segmentation sequence; outputting a predicted label sequence based on the joint probability; and finally training the bidirectional encoding layer and the conditional random field layer with the objective of minimizing the difference between the sample and the predicted label sequence. The technical solution provided by the embodiments of the present application realizes accurate sensitive entity recognition.

The application discloses a network security vulnerability knowledge graph construction method based on HPO-BiLSTM-CRF, which comprises the following steps: collecting public vulnerability data from related databases in the network space security field, preprocessing, and constructing a data set; analyzing and extracting existing data source feature information, and constructing a network security vulnerability domain ontology model CSVDO; based on the optimized bidirectional long short-term memory network and the conditional random field fusion model HPO-BiLSTM-CRF, realizing named entity recognition and relationship extraction; adopting an integrated entity alignment method for knowledge fusion, matching different instances of the same object in different ontologies based on an improved similarity measurement algorithm, and constructing a knowledge graph; carrying out knowledge graph embedding, storing the result into a graph database, and completing knowledge graph construction and graphic visualization. The application can improve the efficiency of entity recognition and relationship extraction in network security vulnerability knowledge, and has the advantages of high efficiency and high accuracy compared with other graph construction methods.

This invention discloses a method, apparatus, device, and storage medium for colorectal gland segmentation. The method achieves accurate segmentation through a three-stage weakly supervised learning process: The first stage utilizes unlabeled colorectal pathological image patches to construct a self-supervised dataset, pre-training and fine-tuning a visual model to obtain a feature encoder adapted to the pathological tissue domain; the second stage constructs an original dataset based on pathological images with image-level gland labels, generating high-quality, refined pseudo-labels through feature extraction, classification network training, attention heatmap generation, image fusion thresholding, and fully connected conditional random field optimization; the third stage constructs a weakly supervised segmentation dataset using these refined pseudo-labels, training a gland segmentation model with a boundary-aware loss function, ultimately achieving pixel-level gland segmentation of unlabeled colorectal pathological images. This invention achieves accurate gland segmentation without pixel-level annotation, effectively reducing annotation costs.

The invention discloses an environmental anomaly broadcasting method and system based on semantic analysis and a knowledge graph, and relates to the technical field of environmental anomaly detection.The technical scheme is characterized by comprising the steps that a current inspection record text is obtained and subjected to semantic analysis, and text keywords are obtained; determining a node corresponding to the text keyword in a preset knowledge graph according to a conditional random field model; the conditional random field model is obtained by training according to a preset constraint matrix; the types of nodes in the knowledge graph comprise enterprises, discharge ports, processes, pollutants and sensitive targets; obtaining abnormal information corresponding to the current inspection record text according to the association relationship between the node corresponding to the text keyword and each type of node in the knowledge graph; according to the method, the matching rule between the text and the node is defined by constructing the constraint matrix, and the conditional random field model is trained according to the matching rule, so that the accuracy of node positioning is improved, and more accurate abnormal information is generated.