4676 results about "Semantic information" patented technology

The invention discloses a hierarchical semantic-driven retrieval enhancement generation method and system, a natural hierarchical relationship and a semantic boundary of a document are effectively reserved by constructing a tree hierarchical structure based on a document chapter title, and a recursive semantic boundary splitting strategy is adopted to refine overlong text nodes, so that the semantic integrity is ensured, and the retrieval enhancement generation efficiency is improved. And the model input length limitation is met, and the semantic information is prevented from being lost. Meanwhile, node knowledge point extraction and abstract generation are achieved through a large language model, top-down multi-level title path transmission and bottom-up content aggregation are combined, the structural perception and semantic expression ability of nodes is enhanced, and in the retrieval stage, based on similarity distribution of query and node semantic expression, an adaptive retrieval threshold value is dynamically calculated, and the retrieval efficiency is improved. A fixed top-k retrieval strategy is replaced, intelligent screening of different query and hierarchical nodes is achieved, information coverage and redundancy suppression are balanced, and retrieval efficiency and accuracy are remarkably improved.

The present invention relates to the technical field of knowledge graph completion tasks, and particularly relates to a knowledge graph link prediction method. The method comprises: using a precoding model to obtain an embedding layer vector, and constructing a corresponding masked triple; adding a corresponding position code to each element in the masked triple, so as to obtain a corresponding input sequence, inputting the input sequence into a trained main masking model, and outputting an entity classification probability; and on the basis of the entity classification probability, predicting potential candidate entities. The method further comprises: concatenating semantic information corresponding to the embedding layer vector and structural information obtained by an embedding model, so as to obtain fused head entity and relation representations, and constructing a corresponding fused masked triple; and adding a corresponding position code to each element in the fused masked triple, so as to obtain a corresponding fused input sequence. The present invention uses a precoding method, thereby effectively reducing the training burden on a model, and improving the inference speed of a model; and a fusion module is used before inputs are fed into a main masked model, thereby ensuring the integrity of textual description information and improving prediction accuracy.

The invention relates to the technical field of data processing, and discloses a cross-border e-commerce commodity recommendation system and method based on multi-source data fusion. The system comprises an acquisition module for performing data acquisition to obtain a cross-border unified data warehouse and federal learning cooperation data; the classification module performs text classification processing to obtain a user preference analysis result and an interpretable attention mark; a quantization module carries out quantization processing to obtain a cross-border selection feature matrix; the fusion module carries out weighted fusion processing to obtain a basic comprehensive score and a weight convergence detection result; the evaluation module performs risk evaluation processing to obtain risk probability distribution and cross-border compliance evaluation results; and the sorting module carries out real-time processing through lightweight preprocessing and a flow-type calculation pipeline to obtain a comprehensive score sorting list of cross-border selected products. The problem that a traditional cross-border e-commerce product selection method cannot effectively integrate multi-source heterogeneous data and cannot analyze user feedback text semantic information is solved.

The invention discloses a cockpit active recommendation system and method based on a knowledge graph and semantic reasoning, and relates to the technical field of intelligent cockpits. The system receives natural language voice input of a user, executes voice recognition and semantic analysis, extracts user intention, keywords and slot entities, generates structured semantic information, constructs or calls a knowledge graph structure with semantic relation edges in combination with environment context information, and obtains the knowledge graph structure with the semantic relation edges. Semantic path reasoning is carried out based on the path dependence weight and the semantic similarity, a semantic edge label guided graph attention mechanism is introduced to calculate a path consistency score, a candidate recommendation set is generated, the semantic fitting degree and the path score are fused to sort and output recommendation content, and the graph edge weight and the user portrait are updated based on user feedback. According to the method, semantic understanding precision, recommendation path interpretability and system adaptive capacity are improved, and the method is suitable for personalized voice recommendation, man-machine interaction and scene linkage control tasks in an intelligent cockpit.

The invention discloses an optical remote sensing image salient target detection method based on progressive attention enhancement, and belongs to the technical field of computer vision. The method comprises the following steps: preprocessing an original data set; inputting the preprocessed image into a hierarchical progressive fusion encoder, capturing a global irregular topological structure and local fine-grained image details, and realizing cross-hierarchical feature fusion; inputting the output characteristics of the encoder into a global context enhancement module, and capturing multi-level context information by adopting a parallel multi-branch structure; and inputting the output features of the hierarchical progressive fusion encoder and the global context enhancement module into a multi-scale progressive attention enhancement decoder, carrying out hierarchical decoding on the input features by adopting a saliency-guided attention mechanism, and gradually aggregating deep semantic information and shallow detail features to realize coarse-to-fine progressive optimization, so as to improve the robustness of the multi-scale progressive attention enhancement decoder. And finally generating a saliency map. The method can effectively improve the processing performance of an irregular topological structure and a complex context relationship in the optical remote sensing image.

The invention relates to the technical field of building information modeling, in particular to a building drawing geometric feature extraction and visual modeling method and system, and the method comprises the steps: carrying out the self-adaptive noise reduction, contrast enhancement and line refinement processing of an original building drawing image, and carrying out the automatic layer separation based on colors and line types; linear geometric features and specific symbol geometric features in the drawing image are extracted, and a geometric feature set is constructed; component instantiation, attribute assignment and topological relation reasoning are carried out by using a predefined building component semantic rule base, and a building component semantic network is generated; and mapping the semantic network to a parameterized three-dimensional modeling engine, calling an IFC standard three-dimensional template, performing parameter driving and automatic assembly, generating a three-dimensional building model with semantic information and a spatial structure, and performing visual output. According to the method, efficient and standardized conversion from a two-dimensional building drawing to a three-dimensional building model can be realized, and the method has the remarkable advantages of processing complex drawings and high-precision modeling.

The invention discloses a voice interaction method and device based on lip language enhancement, equipment and a storage medium, and the method comprises the steps: extracting lip language features based on an image sequence of a lip region, and carrying out the feature extraction of a voice signal, and obtaining an audio feature; performing cross-modal fusion coding on the lip language features and the audio features to generate mixed features containing audio-visual information; inputting the mixed features into a large language model, understanding the intention of the interaction object and generating a corresponding semantic reply; and finally, synthesizing into voice and / or converting into characters. According to the invention, by introducing the lip features, additional visual clues are provided for speech recognition, and the robustness and accuracy of speech recognition can be significantly improved; effective fusion coding is carried out on the lip language features and the sound features, and semantic information splitting caused by simple and independent recognition is avoided; and the capability of the large model is fully utilized, so that more natural and more intelligent interaction experience is realized.

The invention discloses a multi-modal fusion rumor detection method and system based on a dynamic graph convolutional neural network. According to the method, a dynamic feature graph of a language propagation path is constructed, and potential features in the language propagation process are extracted and analyzed by utilizing time sequence changes and key node relations between nodes in a propagation graph. A neural network is adopted to extract and enhance image data, text semantic features are extracted in combination with a text feature modeling network, text feature vectorization expression is achieved based on a BERT model, and rich semantic information is obtained. And a gating mechanism is introduced to dynamically adjust fusion weights of different modal features, and an information fusion strategy is optimized. A collaborative attention mechanism is further adopted for deep fusion, interactive learning of text, image and propagation path features is enhanced, and the relevance of cross-modal and time series data is improved. And finally, inputting the fused feature vectors into a classifier for accurate classification, thereby realizing accurate detection of the social media rumors. According to the method, the multi-modal features are effectively integrated, and the false information identification efficiency is remarkably improved.

The invention discloses a virtual-real combination-based power generation side industrial control system network security target construction method. The method comprises the following steps of: constructing a virtual-real combination target environment consisting of a physical equipment layer and a virtual model layer; the heterogeneous industrial control protocol between the physical equipment layer and the virtual model layer is analyzed, protocol semantic information is extracted, and a bidirectional dynamic mapping rule of a physical equipment state and a virtual model state is generated based on the protocol semantic information; based on a state change event of the physical equipment layer, according to a bidirectional dynamic mapping rule, synchronizing changed equipment state data to the virtual model layer in real time, and simulating protocol behavior logic corresponding to the equipment state data in the virtual model layer according to a security test requirement; and based on the attack instruction or the abnormal state signal generated by the virtual model layer, according to the protocol specification format of the target physical equipment, converting the instruction or the signal into an executable control command, and driving the physical equipment layer to execute an operation corresponding to the control command.

The invention discloses a visual inertial positioning method based on dynamic target detection and semantic information constraint, and belongs to the field of motion estimation and dynamic environment processing. According to the method, a dynamic target detection mechanism is introduced, the dynamic target is effectively detected based on the target detection network, inertial navigation information and geometric constraints, the dynamic target is effectively recognized in the image processing process, the corresponding dynamic feature points are screened out, the mismatching rate of the dynamic features is remarkably reduced, and high-quality observation input is provided for back-end optimization. In the back-end sliding window optimization stage, a semantic information consistency constraint method is constructed, and the estimation stability of the system in a weak texture area or a repeated texture area is enhanced by utilizing the consistency of feature points in the same semantic area on a geometric structure. Visual inertia pose estimation is realized based on dynamic target detection and semantic information constraint, and high-robustness and high-precision pose estimation can still be realized in a complex environment with dynamic interference of pedestrians, vehicles and the like and severe scene change.

The invention discloses an unmanned aerial vehicle aerial image target detection method based on PSO-DETR, and belongs to the technical field of unmanned aerial vehicle aerial image target detection. Firstly, a parallel patch perception attention feature extraction module is constructed, and an efficient multi-branch backbone network C3KCSPnet is designed by fusing a CSPDarknet53 structure; according to the network, gradient flow is improved through deep optimization, and the capturing capacity of high-level semantic information is enhanced. And secondly, an enhanced channel offset hybrid operator is provided, the dependency relationship between channels is enhanced through a channel shuffling mechanism, and cross-channel interaction of local space information is realized in combination with channel offset operation, so that the feature recovery quality and fusion efficiency in an up-sampling stage are improved, and the problem of missed detection of a shielded target is further relieved. And finally, a re-parameterization hierarchical aggregation network is designed, effective integration of shallow details and deep semantics is realized through an efficient hierarchical fusion mechanism on the premise of ensuring controllable calculation complexity, and the detection performance of the small target is further enhanced.

The invention discloses a control method and equipment for an intelligent robot with a body and a storage medium, and belongs to the technical field of robots. The method comprises the steps of receiving a task instruction and environment perception data, performing fusion processing on the task instruction and the environment perception data, generating task semantic information associated with the task instruction, analyzing the task semantic information through an implicit planner, generating an implicit action mark, and based on the implicit action mark, generating an implicit action. And decomposing a to-be-executed task corresponding to the task instruction into a plurality of task sub-targets layer by layer, generating an action instruction sequence based on the task sub-targets, and executing a control action corresponding to the action instruction sequence. According to the method, based on the constraint of the implicit action mark and the action instruction sequence, the robot with the body can flexibly respond to environment parameter changes such as object position deviation or new task requirements in the task execution process, the coupling degree of the action instruction of the robot with the body and the predefined scene is reduced, and high stability of task execution is ensured.

The invention relates to the technical field of data processing, and particularly provides an operation intention recognition method, system and device based on multi-mode fusion and a medium, and the method comprises the steps: synchronously collecting interaction data of at least two modes of a user, the modes comprising at least two of gestures, voice and eye gaze; carrying out alignment processing on the interaction data, wherein the alignment processing comprises time synchronization and space mapping to a unified coordinate system; recognizing structured semantic information from each piece of aligned modal data, wherein the structured semantic information comprises a gesture type, a voice text and a fixation point coordinate; and based on a preset semantic rule and context memory, performing semantic association and anaphora resolution on the structured semantic information to obtain an operation intention. The method effectively overcomes the inherent defects of unnatural single-mode interaction, easy ambiguity and poor fault tolerance.

The invention discloses a time sequence knowledge graph reasoning method based on a large language model, which comprises the following steps of: performing vectorization expression on query entities and relationships by utilizing an embedded model, and generating vectors containing structural features and semantic information in combination with adjacent entities, relationships and time information; a candidate entity set most relevant to query is screened out from the large-scale entity set by calculating the conditional probability; constructing an event evolution tree, and performing multi-hop sampling on adjacent nodes of a query entity and a candidate entity in a time window to generate a tree structure capable of completely reflecting a historical event evolution path; the method comprises the following steps: designing a structured prompt template, and performing instruction fine tuning on a pre-trained large language model by adopting an LoRA (Low Rank fine tuning technology), so that the model better follows a task instruction to generate a reasoning result. And efficient and accurate time sequence knowledge graph reasoning is realized.

The invention belongs to the technical field of vehicle trajectory prediction, and discloses a multi-scale interactive perception highway vehicle trajectory prediction method, which comprises the following steps: jointly modeling short-term burst features and long-term evolution trends through a convolutional neural network and a bidirectional gating cycle unit, and introducing a time sequence attention mechanism to improve the perception ability for key time slices; in combination with a dynamic graph attention mechanism including physical edge features such as relative position, relative speed and relative acceleration, a vehicle interaction relationship is updated in real time so as to improve spatial modeling precision and interpretability; in the decoding stage, the guide vector and the semantic information of the lane are fused, so that the predicted trajectory conforms to the geometric structure of the road in space and keeps smooth and continuous in time. According to the method, the robustness and adaptability of the model in the sparse adjacent vehicle environment of the expressway can be improved while the prediction precision is ensured, a more stable and reliable trajectory prediction result is provided for an intelligent traffic system, and powerful technical support is provided for traffic safety management and operation scheduling of the expressway.

The invention relates to the technical field of power grid monitoring, in particular to a power operation risk identification method, system and device based on multi-modal data fusion and a storage medium. In order to solve the problems of multi-modal data splitting, topological constraint missing and the like in traditional power disturbance analysis, an improved BERT model is constructed, and electrical signal time-frequency features and text semantic information are mapped to a unified vector space through a multi-modal embedding mechanism; a time sequence attention mechanism is adopted to establish a time dependency relationship between signals and texts, and a graph attention network is combined to realize risk propagation modeling under power grid topology constraints; and collaborative optimization of disturbance classification, risk prediction and trend analysis is carried out through a multi-task learning framework. The technical problems that heterogeneous data fusion is difficult and risk identification precision is insufficient are effectively solved, accurate identification and intelligent early warning of electric power operation risks are achieved, and the safe operation level of a power grid is improved.

The invention discloses a brain tumor multi-mode large model construction method, device and equipment and a storage medium, and is applied to the technical field of brain tumor imagines.The method comprises the steps that pixel-concept level alignment is conducted on a multi-mode MRI image and a pathological text; constructing a multi-modal feature fusion network for fusing image features and text features by adopting an attention mechanism of pathology perception and combining medical semantic information; training the multi-modal feature fusion network to generate an analysis report and a segmentation result; according to the technical scheme of multi-task cooperation, cross-modal pathological semantic accurate alignment, pathological knowledge graph injection and lightweight and continuous optimization parallelization, full-process coverage of brain tumor accurate segmentation, analysis report generation and prognosis prediction is achieved, the problems that a traditional model lacks pathological semantic support and is insufficient in clinical adaptability are solved, and the clinical adaptability of the traditional model is improved. And the deployment feasibility and the dynamic optimization capability are also considered.

The invention relates to the technical field of intelligent driving, in particular to a method for generating an obstacle avoidance and planning cooperative path in an urban complex environment, which comprises the following steps: acquiring environment real-time sensing data through a vehicle-mounted multi-sensor, and constructing a dynamic semantic traffic matrix in combination with high-precision map static semantic information; based on the matrix, a multi-objective optimization algorithm is adopted to calculate the security cost, the efficiency cost and the rule conformity cost of the path, and a global optimization path is generated; inputting the global optimization path and the dynamic obstacle motion vector into an intention prediction model, generating dynamic obstacle future trajectory probability distribution and interactive intention classification, and further generating an avoidance strategy and adjusting a local path in real time; and inputting the adjusted local path into a kinematics model to carry out kinematics feasibility verification, and outputting an executable track or path re-planning. According to the method, the integration of dynamic environment understanding, path planning and obstacle avoidance strategies is realized, and the method is suitable for the path generation task of an automatic driving system in an urban complex traffic scene.

The invention discloses an electric power scene defect small target detection method based on Gaussian mask supervision and cross-layer attention guidance, and the method comprises the steps: inputting an electric power scene image into a detection model, extracting an initial feature map through a backbone network, carrying out the multi-stage feature extraction of the initial feature map according to a convolution path, and carrying out the multi-stage feature extraction of the initial feature map; processing the multi-stage features based on a path aggregation network, and outputting a plurality of fusion feature maps with different feature levels from shallow to deep; and based on cross-scale window attention, guiding a shallow fusion feature map to carry out semantic information modeling by using a deep fusion feature map with high semantics in every two adjacent fusion feature maps, and after a plurality of output feature maps are obtained, respectively processing and outputting prediction results by using a multi-branch detection head. According to the method, shallow feature activation prediction and cross-scale window attention guidance are fused, and the detection robustness and positioning precision of a tiny fault target in an unmanned aerial vehicle inspection image can be effectively improved.

The invention discloses a recommendation method for enhancing semantics and interest perception by using a large language model. The method comprises the following steps: firstly, performing semantic modeling on unstructured text information such as user comments, article description and the like by utilizing the powerful capability of a large language model in semantic comprehension and user preference modeling aspects, so as to improve the deep perception capability of a recommendation system on user interests and article semantic attributes; then, through a semantic feature alignment and discretization strategy, the problem that continuous semantic representation generated by a large language model is incompatible with features of a traditional recommendation system in the aspect of an expression structure is solved; finally, unified modeling of semantic information and traditional recommendation signals is achieved through a recommendation integration mechanism, and recommendation performance and model interpretability are improved.

The invention provides a remote sensing image semantic segmentation method and system fusing a convolutional neural network and a visual state space model, and the method comprises the steps: firstly extracting the multi-scale semantic features of a remote sensing image based on a lightweight ResNet18 encoder; secondly, a decoder based on a visual state space module is used for modeling a long-distance dependency relationship in an image and recovering spatial resolution; the local feature compensation module is used for enhancing the perception capability of fine-grained semantic information and improving the segmentation precision of a small target area; and finally, the multi-scale attention enhancement module is used for fusing deep and shallow layer features and realizing collaborative optimization of spatial details and semantic information. According to the method, the global semantic features and the local detail information of the remote sensing image can be extracted at the same time, and the method has high segmentation precision and a good remote sensing image semantic segmentation effect.

The invention discloses a multi-modal index knowledge base and a construction method thereof. The construction method comprises the following steps: processing a heterogeneous document to obtain a semi-structured document; identifying a title hierarchical relationship of the document to construct a document logic structure; carrying out minimum chapter blocking on the text of the semi-structured document to obtain logic blocks; performing semantic segmentation on each logic block to obtain text blocks; the method comprises the following steps: constructing text block nodes by meta-information of text blocks, constructing non-text block nodes by meta-information of non-text elements, extracting document nodes, chapter nodes and chapter-chapter inclusion relationships according to a document logic structure, respectively extracting semantic information from the text blocks and the non-text elements, and storing the semantic information in a database; recording the corresponding relationship between the text block nodes and the semantic information and between the non-text block nodes and the semantic information; constructing a knowledge graph based on each node and relationship and storing the knowledge graph into a graph database; and constructing semantic knowledge based on the semantic information and storing the semantic knowledge into a vector database. According to the scheme, lossless retention of multi-modal information and structured organization of document logic are realized, and efficient indexing and accurate recall are facilitated.

The invention discloses a light-weight instrument small target detection model and method for a complex industrial scene, belongs to the crossing field of deep learning and edge calculation, and aims to solve the problem that an existing method cannot meet the real-time detection requirements of edge equipment such as an inspection robot in the aspects of precision, efficiency and small target detection capability. The model comprises a lightweight backbone network used for extracting multi-scale features from an input image; the cross-scale feature fusion network is used for bidirectionally fusing the multi-scale features, retaining shallow space details and deep semantic information and outputting fused features; the deformable large-kernel target sensing module is deployed at a specified position of the cross-scale feature fusion network so as to better capture feature information related to a target area and improve the feature expression capability of a small target; and the multi-task decoupling prediction network performs classification, positioning regression and confidence prediction on the input image in parallel, and a positioning regression branch adopts an EIOU loss function of decoupling width and height optimization.

A clinical vertebra image segmentation method for assisting pedicle screw placement surgery, said method comprising: constructing a VerseDiff-UNet end-to-end framework, the framework being integrated with a denoising diffusion probabilistic model (DDPM); combining a noise-added image with a marked mask by using the VerseDiff-UNet framework, and guiding a diffusion direction toward a target region; and introducing a shape priors module on the basis of the DDPM, and extracting structural semantic information from an input spine image. In order to capture specific anatomical prior information in a medical image, the shape priors module is combined and the module effectively extracts the structural semantic information from the input spine image, thereby enabling more accurate anatomical structure segmentation, and facilitating accurate diagnosis and treatment of spinal disorders.

The invention belongs to the technical field of bridge health monitoring, and discloses a multi-modal fusion bridge disease detection and three-dimensional point cloud registration method, which specifically comprises the following steps: S1, constructing a bridge disease data set by adopting four public data sets CFD, CrackTree200, Crack500 and CrackSeg9k and autonomously acquired data samples; a bridge disease semantic segmentation result is introduced as a semantic clue, a multi-level semantic consistency registration frame is constructed, intra-class mismatching is inhibited in combination with a scene semantic consistency mask matching module, and the three-dimensional point cloud registration precision is improved; a multi-radius annular region semantic feature extraction and clustering optimization strategy is adopted, the feature integrity of sparse point cloud data is enhanced, RGB-D multi-scale features are fused based on a lightweight encoder-decoder framework, accurate segmentation of complex diseases is realized through a double-branch attention module and an adaptive pyramid context module, and the accuracy of segmentation of the complex diseases is improved. And semantic information is deeply fused with the three-dimensional point cloud to form closed-loop feedback.

The invention relates to the technical field of panoramic image construction, in particular to a panoramic image reconstruction method and system based on multi-angle imaging. The method comprises the following steps: collecting a multi-angle original image based on a distributed multi-camera array, carrying out adaptive filtering denoising and adaptive panoramic imaging adjustment, and constructing a multi-angle imaging geometric constraint network; performing multi-view semantic information deviation elimination based on a multi-angle imaging geometric constraint network, and performing global semantic feature fusion to obtain a unified semantic space representation framework; identifying illumination feature information of different visual angles, performing multi-angle illumination corresponding compensation on the multi-angle original image, performing image semantic distortion correction based on a unified semantic space representation framework, and constructing a multi-angle illumination compensation image; and performing multi-scale texture structure analysis on the multi-angle illumination compensation image to generate a high-fidelity texture fusion image. According to the invention, a natural and seamless panoramic image is provided, a panoramic scene is perfectly presented, and the immersive visual experience of a user is improved.

The invention discloses a cerebral apoplexy risk and prognosis prediction system and method, and relates to the field of intelligent medical treatment, and the system comprises a data processing and knowledge construction layer which is used for extracting, cleaning and constructing a space-time multi-modal knowledge graph and structured clinical features from multi-source heterogeneous medical data; the feature engineering and fusion layer is used for deeply fusing dynamic semantic information in the space-time multi-modal knowledge graph and the structured clinical features through a graph embedding and attention mechanism to generate a fusion feature vector for a cerebral apoplexy prediction task; and the prediction model and output layer is used for performing cerebral apoplexy risk and prognosis prediction based on the fusion feature vector to obtain a prediction result, and generating a decision result for assisting a doctor in understanding the model through an interpretable mechanism. The method provided by the invention can improve the accuracy of stroke recurrence, bleeding transformation or function prognosis prediction, and provides a new way for accurate stroke management.

The invention discloses a sparse view angle three-dimensional reconstruction method based on 3DGS. The method comprises the following steps: providing a Gaussian field initialization model; according to the method, the FPN, the 2D U-Net and the MLP are used for forming a Gaussian parameter initialization network, the network integrates high-level and low-level semantic information, and the understanding of the network on multi-scale features is enhanced. And then a complete Gaussian field model is obtained by combining initial point cloud coordinates generated by DUSt3R. And then performing projection and rasterization under different visual angles on the 3D Gaussian ball by taking three-dimensional Gaussian splashing as a main body frame of an algorithm, performing adaptive adjustment of cloning and pruning on the 3D Gaussian ball with relatively large gradient change, and outputting an output result which is optical information and depth information of each visual angle image after rendering. Then, in gradient back propagation, pixel value absolute value deviation between the RGB image obtained by rendering and the RGB true value is supervised through L1 norm loss; the scene rendering quality is high, calculation is simple, and engineering implementation is easy.

The invention discloses a coal bunker reserve real-time monitoring method integrating a laser radar and a camera. The method comprises the following steps: step 1, carrying out space-time joint calibration and hardware-level synchronization on a laser radar and a camera, and establishing a projection relationship between a point cloud and an image; 2, cooperatively collecting coal bunker data based on the calibration parameters, and obtaining a point cloud and a high-resolution image; step 3, heterogeneous preprocessing is carried out on the collected point cloud and the image, denoising, segmentation and down-sampling are carried out on the point cloud, semantic segmentation is carried out on the image, and coal and foreign matters are identified; step 4, based on image texture and point cloud coordinate coupling mapping, reconstructing a three-dimensional model with a semantic tag, projecting the point cloud to an image plane to perform feature matching so as to endow the texture, and fusing image semantic information to correct the model; and 5, dynamically calculating the volume of the coal pile by using the semantic three-dimensional model to realize real-time monitoring of reserves. The system overcomes the limitation of a single sensor, achieves the real-time and high-precision monitoring of the reserves of the coal bunker, and effectively improves the intelligent level of coal bunker management.

The invention provides a depth map generation method and device based on a large model, a three-dimensional reconstruction method and device, electronic equipment and a storage medium, relates to the technical field of artificial intelligence, in particular to the technical fields of computer vision, deep learning, large models and the like, can be applied to real-time road scene depth perception, environment three-dimensional reconstruction and obstacle avoidance, and can be applied to real-time road scene depth perception. And virtual and real scene fusion and other scenes can be realized. The specific implementation scheme is as follows: performing visual coding on a monocular image to obtain a coded image; inputting the coded image and the target text into a pre-trained large language model for fusion to obtain fusion features; generating global guide features based on the fusion features, wherein the global guide features comprise joint semantic information of visual features and text features; adding noise to the color image of the monocular image to obtain a noise feature sequence; de-noising the noise feature sequence under the condition of the global guide feature, and generating an implicit feature matched with the joint semantic information; a depth map is generated based on the implicit features.