120 results about "Information quality" patented technology

The invention relates to the technical field of robot decision making, and provides a robot multi-modal fusion autonomous decision making method and system based on a large language model.The method comprises the steps that a robot obtains multi-modal environment information through a visual sensor, a touch sensor, an auditory sensor and a laser radar which are carried by the robot; performing preliminary filtering and noise reduction processing on the original sensor data, and synchronously recording all the sensor data by timestamps; performing space-time semantic alignment on the preprocessed multi-modal data, mapping pixel coordinates of a target in a visual target coordinate quantization original image to a robot coordinate system, performing uncertainty evaluation on a multi-modal signal through a dynamic Bayesian network, and taking entropy or variance as an uncertainty quantitative evaluation index. According to the method, the information quality is improved from a data fusion source, accurate and reliable basic support is provided for subsequent decision making, and decision making errors caused by data deviation are greatly reduced.

The invention discloses an intelligent enterprise financial data analysis system and method based on artificial intelligence, and belongs to the technical field of enterprise informatization management and intelligent finance. The objective of the invention is to solve the technical problems of low processing efficiency, insufficient data insight, lagging risk identification, weak intelligent decision-making assistance capability and the like in the existing enterprise financial data analysis process. The system generally integrates a data automatic acquisition and integration module, a financial data deep analysis and modeling engine based on artificial intelligence, a key financial index and operation performance intelligent calculation module, a multi-dimensional financial risk intelligent identification and dynamic early warning module and an interactive visual analysis report and decision support module. The core of the method is that internal and external multi-source heterogeneous financial and operation data of an enterprise are automatically collected and fused, deep processing and intelligent analysis are performed on the data by applying artificial intelligence technologies such as machine learning and natural language processing, and accurate portrait of the financial condition of the enterprise, dynamic evaluation of operation performance and real-time monitoring and prediction of financial risks are realized. And analysis reports and optimization suggestions with insight are generated, so that enterprise managers can make efficient and scientific operation decisions. According to the method, the automation and intelligence level of enterprise financial data analysis can be remarkably improved, the financial information quality and decision support efficiency are improved, the risk prevention capability of enterprises is enhanced, and the enterprises are assisted to realize refined operation and value creation.

The invention relates to the technical field of food safety intelligent tracing, and discloses a livestock and poultry product epidemic disease and tracing information quality tracing system based on artificial intelligence. The system comprises a raw material file construction module, an abnormal active prediction module, a multi-modal information fusion module, a tracing knowledge graph construction module and an intelligent tracing decision module. The system generates a raw material archive by collecting livestock and poultry raw material information, and analyzes production monitoring data in real time by using a time sequence model to actively predict abnormal nodes of epidemic diseases. After the abnormity is recognized, multi-link heterogeneous traceability information is synchronously called and fused, an abnormal event panorama is generated, and then a traceability knowledge graph is dynamically constructed in combination with an external knowledge base. And finally, automatically generating an optimized traceability query path according to the knowledge graph, and driving a precise traceability process. According to the invention, the conversion from passive recording to active early warning is realized, and the accuracy and efficiency of epidemic disease tracing are improved through deep fusion and intelligent reasoning of multi-source information.

The invention provides an unstructured data compression method and device, equipment and a storage medium, and the method comprises the steps: carrying out the feature extraction and clustering processing of a to-be-compressed unstructured data set, and obtaining a data segment clustering group; performing semantic region segmentation processing on each data fragment in the data fragment clustering group to obtain a semantic important region and a semantic non-important region; performing differential coding processing on the semantic non-important regions in the same clustering group to obtain differential coding data; and performing high-quality compression processing on the semantic important regions in the same clustering group, and performing merging processing on the compressed important semantic data and the difference coding data to obtain an unstructured data compression result. According to the method, semantic region segmentation and differential compression processing are performed on the unstructured data, so that the data compression efficiency can be effectively improved on the premise of ensuring the quality of important information, and an efficient solution is provided for storage and transmission of large-scale unstructured data.

The invention relates to the technical field of data processing, in particular to a laser and binocular camera-based position positioning system, which comprises an acquisition unit, a processing unit, a fusion unit, an abnormality judgment unit, an adjustment unit and a correction unit. According to the method, multi-source information such as a laser point cloud matching result, binocular visual odometer output and the number of feature points is subjected to joint modeling in an extended Kalman filter, so that the AGV obtains stable pose estimation in different environments, and the pose estimation accuracy is improved by analyzing the change trend of pose covariance in an adjusted monitoring window. A density threshold value, an information entropy threshold value and a dynamic proportion threshold value are automatically corrected, so that the threshold values can adapt to real working conditions for a long time, and performance degradation caused by a fixed threshold value is avoided; the problems that due to the fact that sensor information quality fluctuation cannot be recognized and processed, positioning precision is lowered, accumulative errors are increased, and positioning loss is prone to occurring in a complex environment are effectively solved.

The invention relates to the field of robot navigation, and particularly discloses a spatial perception observation information quality evaluation method and system, a terminal and a medium, and the method comprises the steps: converting the measurement information of each dimension into a normalization probability according to the measurement target of each dimension; according to the concept of entropy in the information theory, the normalized probability of each dimension is converted into a computable entropy value; a downstream task weight coefficient is introduced for the entropy value of each dimension, the multi-dimensional entropy value after the downstream task weight coefficient is introduced is mapped into a physical interpretable confidence coefficient in a [0, 1] interval through an exponential attenuation function, and the confidence coefficient explicitly represents the uncertainty of observation information of each dimension. According to the method, a space perception quality evaluation system with mathematics interpretability, scene adaptation and real-time calculation is formed, multi-source uncertainty is mapped into executable confidence, and multi-dimensional accurate evaluation of intelligent agent space perception observation information is achieved.

The invention relates to the field of data analysis and cleaning, in particular to a method and system for automatically collecting and cleaning public opinion experimental data. The method comprises the following steps of performing periodic data crawling on a public information website, performing public opinion hotspot analysis, and extracting a to-be-processed hotspot event stream; performing keyword detection and global content distribution analysis on the to-be-processed hot event stream to obtain a keyword frequency distribution diagram; performing negative emotion deep recognition according to the to-be-processed hot event stream, and extracting negative emotion public opinion data; performing negative public opinion evolution trend prediction based on the keyword frequency distribution diagram to obtain a negative public opinion propagation situation; and performing IP address tracing on the negative emotion public opinion data to obtain the negative public opinion region sensitivity. According to the method, comprehensive and accurate abnormal public opinion detection and deletion are realized, and the network information quality and the public governance level are improved.

The invention discloses an image information quality control method and system, an automatic detection platform is realized by establishing a set of standardized image quality verification process and technical specification, a core module carries out batch technical index verification and abnormal image data detection, a first analysis sub-module carries out content integrity verification, and a second analysis sub-module carries out content integrity verification. The content reconstruction sub-module carries out content reconstruction, the second analysis sub-module carries out format normalization detection, and the format restoration sub-module finally carries out format restoration, so that multi-level quality verification is realized through an automatic platform, and the automatic detection platform is adopted to carry out batch technical index verification on original image data. According to the method, content integrity checking and format normalization detection are performed on the abnormal image, and data reconstruction processing is performed, so that automatic detection, restoration and continuous optimization of image quality are realized, image processing efficiency and quality control level are improved, and the method is suitable for quality assurance requirements of various image data.

The invention relates to the technical field of remote sensing geological information processing, and discloses a mineral geological exploration system based on remote sensing image texture analysis, which comprises a time sequence base line library construction unit, an ecological proxy texture channel processing module, a remote sensing image texture analysis unit and a remote sensing image texture analysis unit, the system comprises a normalization vegetation index texture purification module for purifying a normalization vegetation index texture of a current image according to a time sequence statistical baseline so as to generate an ecological proxy texture anomaly graph, a texture channel construction processing module for generating a texture anomaly graph, and a collaborative verification engine module for performing collaborative verification on the texture anomaly graph before executing spatial coupling judgment. According to the method, the information quality of two channels is evaluated according to image information entropy, verification logic is dynamically selected, earth surface covering information regarded as interference in traditional exploration is converted into an independent verification dimension, and an evidence chain of internal cross verification is constructed through double constraints of space and time and self-adaptive evaluation of the information quality.

The invention is suitable for the technical field of mixed modal retrieval, provides a financial risk identification method based on multi-granularity mixed modal retrieval, and is suitable for the field of mixed modal retrieval. The method comprises the following steps: constructing a multi-modal retrieval library, respectively encoding image and text modal information by using CLIP and RoBERTa models, and aligning features through a learnable projection matrix to form a searchable financial knowledge base; a multi-granularity retrieval scheme combining Dual-Encoder and Cross-Encoder is adopted, a Top10 candidate answer is rapidly positioned through coarse granularity, and a high-relevance candidate set is accurately screened through fine granularity; and finally, through a multi-input weighted integration strategy, integrating multiple rounds of prediction results of the large model and carrying out weighted summation based on association degrees to generate a final risk prediction result. According to the method, limitation of a single data source is broken through, comprehensive utilization of cross-modal information is realized, information quality and retrieval efficiency are ensured, model uncertainty is remarkably reduced, a risk identification system with robustness and accuracy is formed, and scientific decision support is provided for financial institutions.

The invention relates to the technical field of computer information processing, and provides an early warning information quality inspection method and device based on an AI large model, and a medium. The method comprises the following steps: inputting an early warning service information text to be subjected to quality inspection, which is screened out according to an early warning trigger source identifier and an early warning level label carried by a front push optimization instruction, into a pre-trained AI large model, and executing semantic analysis mining in combination with a preset early warning field knowledge base, outputting the core semantic elements, the potential association risk features and the logic dependency relationship; generating a dynamic semantic error correction rule base based on semantic association mapping of the core semantic elements and the potential association risk features; using a cross-modal semantic understanding layer of the AI large model to execute semantic deviation recognition and correction guide processing to obtain a semantic problem recognition result; generating early warning correction information for balancing semantic accuracy and risk conductivity through a semantic reconstruction layer; and the push strategy generation network driven by the AI large model is input to generate a differentiated front push priority sequence.

The invention belongs to the technical field of geographic information processing, and discloses a CORS surveying and mapping information quality control method combined with space-time big data analysis. The method comprises the following steps: identifying an ionized layer abnormal disturbance region based on spatio-temporal multi-scale wavelet decomposition, accurately positioning a resolution attenuation boundary through a second derivative mutation feature, constructing a resolution defect compensation weight matrix, and generating an adaptive enhanced high-resolution delay prediction field by using tensor decomposition. Empirical mode decomposition is adopted to recognize a weak fluctuation mode, early capture of precision degradation threatening features is achieved, a time-varying quality threshold curved surface is dynamically constructed based on an evolution trend, and grading quality marking of CORS observation data is achieved. According to the invention, the misjudgment rate of quality evaluation is reduced, and the reliability of surveying and mapping information is improved.

The invention relates to the technical field of insurance business processing, and particularly discloses an automatic car insurance business handling method based on an AI customer service robot. The method comprises the following steps: receiving and standardizing business related information through an information acquisition unit, screening effective information through an information quality evaluation model, extracting key fields by means of OCR text recognition, and generating structured insurance data in combination with multi-modal consistency comparison; a vehicle information system is called to verify data integrity and accuracy in real time and automatically complement and correct errors, and compliance verification is completed through an underwriting rule engine according to a preset rule of an insurance company to generate standardized insurance data; a rate engine is called based on the data to generate a quotation scheme, the quotation scheme is submitted to an underwriting system to generate an underwriting conclusion after being confirmed by the user, a payment two-dimensional code is automatically generated after underwriting is passed, an electronic insurance policy is pushed when the user pays successfully, and meanwhile, whole-process data is stored persistently, and a service ledger is generated.

The invention discloses an unmanned aerial vehicle autonomous strike decision self-adaptive adjustment system based on multi-source information fusion, and particularly relates to the technical field of unmanned aerial vehicle autonomous control and self-adaptive decision, the system comprises an information fusion and decision inner ring and a meta-cognitive self-adaptive adjustment outer ring, the inner ring executes multi-source information fusion and generates a strike decision, and the outer ring executes a meta-cognitive self-adaptive adjustment. And the outer ring evaluates the environment and information quality in real time, drives the adjustment strategy generator based on reinforcement learning by comparing the expected efficiency and the actual efficiency of the decision, and dynamically adjusts the fusion weight and the decision parameter of the inner ring or switches the decision model. The problems that an existing system is poor in environmental adaptability and cannot achieve online self-optimization are solved, autonomous, closed-loop and continuous optimization of decision-making performance in a dynamic battlefield environment is achieved, and the autonomous combat effectiveness, the environmental adaptability and the task robustness of the unmanned aerial vehicle are remarkably improved.

The invention discloses an intelligent mobile terminal tourism information personalized pushing system based on a user portrait, and relates to the technical field of mobile terminal tourism pushing, and the system comprises a user portrait construction module which collects basic information, mines behavior data and generates an interest label; the tourism information processing module is used for collecting multi-source tourism data, carrying out structured processing and evaluating the quality; the personalized push decision module is used for matching user interests and situations and determining a push strategy; the interaction feedback module is used for tracking user operation and optimizing portraits; and the privacy protection module is used for desensitizing the data, managing the authority, encrypting and storing. According to the method, the pushing accuracy and timeliness are improved, the user experience is optimized, the information quality and privacy security are guaranteed, and the method is adaptive to the whole travel process demand.

The invention discloses a multi-modal heterogeneous knowledge base automatic construction and evolution updating method and system, and belongs to the technical field of artificial intelligence and knowledge base data processing, and the method comprises the following steps: S1, multi-modal data collection and cross-modal semantic alignment; s2, human-social multi-source data labeling and information quality screening are carried out; s3, constructing and dynamically updating a multi-modal heterogeneous knowledge base; and S4, system integration and packaging service: packaging the above processes into an integrated multi-mode knowledge base construction and updating system. Unified representation, high-quality knowledge precipitation and dynamic evolution of multi-source heterogeneous data in the human society field can be realized, and a high-quality, reasonable and high-timeliness knowledge infrastructure is provided for human society intelligent service.

The invention provides an insurance product design method and device, electronic equipment, a chip and a storage medium, and the method comprises the steps: obtaining the static information and behavior information of at least one user, and enabling the behavior information to be obtained through employing a first network to carry out the feature extraction of the operation of the at least one user; inputting the static information and the behavior information into a trained insurance business reasoning module to obtain structured data, the structured data being used for indicating a corresponding relationship among the static information, the behavior information, the insurance industry information and the risk information; and generating a target insurance product according to the structured data and the generation target. Insurance industry information, risk information and static information and behavior information of a user can be analyzed and sorted to determine a corresponding relationship between requirements and characteristics of the user in the insurance industry and information quality inspection related to an insurance product, and then a target insurance product can be generated according to structured data. The generated insurance product can better meet user requirements, and the risk is lower.

The invention discloses an intelligent management and control method based on the full life cycle of a steel structure, and relates to the technical field of intelligent management of steel structures. According to the method, the bridge steel structure digital intelligent manufacturing cloud platform is constructed, the steel structure full-life-cycle intelligent control method is designed based on the cloud platform, the component production technological process is randomly set through the graphical process designer and the cloud platform, information is automatically circulated in the component manufacturing process, project component data is clear and visible, and the construction efficiency is improved. The production management level and production benefits are improved, the technological process is optimized through the cloud platform, production tasks are distributed, teams and related production personnel receive the tasks through mobile phone APPs, production entry information, quality information and the like are filled in the production process, production component models and part models are visually browsed in the production process through the BIM technology, and the production efficiency is improved. The project condition is visually browsed on the project site by means of the BIM technology, and the installation progress of the project is mastered.

The invention discloses a text and image multi-mode user interest identification method and system oriented to private domain communities. The system comprises a text feature extraction module, an image feature extraction module, a multi-modal fusion module, an interest probability aggregation module and a dynamic threshold judgment module. The multi-modal fusion module adopts a gating mechanism to realize self-adaptive fusion, a gating factor is automatically adjusted according to the characteristics of an input sample, and weighting is carried out between a text mode and an image mode, so that the robustness of a fusion result can still be ensured when the quality of information in different modes is unbalanced. And the interest probability aggregation module performs time sequence smoothing on the instant interest probability through an exponential weighted moving average method, and inhibits short-time noise interference. And the dynamic threshold value judgment module is used for calculating a dynamic threshold value according to the mean value and the standard deviation of the recent window, comparing the smoothed interest probability with the dynamic threshold value, and judging that the corresponding interest label is activated when the smoothed interest probability is greater than or equal to the dynamic threshold value. According to the method, the user interests can be accurately recognized in a complex and changeable private domain community environment, and the adaptability, stability and real-time performance of interest recognition are achieved.

A method for compressing Sounding Reference Signal (SRS) channel information in massive MIMO systems is disclosed. When SRS channel estimation is performed in a Radio Unit (RU), the resulting channel estimates create significant fronthaul bandwidth demands. The method transforms SRS channel estimates from antenna space to domains with superior energy compacting properties, enabling transmission of only a subset of channel values while preserving information quality. The method includes beam space compression using mathematical transforms such as DFT, SVD, or EVD to convert channel estimates to beam space, followed by selection of the most significant beams. Alternative embodiments include tap domain compression using DCT or DFT, and delay-Doppler domain compression. The methods support static beam space compression using predefined beam bases and dynamic compression employing adaptive beam bases. O-RAN C-Plane implementations extend Section Type 6 to support the compression methods. The methods significantly reduce fronthaul bandwidth requirements while maintaining channel estimation accuracy, enabling improved system capacity in massive MIMO deployments.

This invention relates to the field of battery parameter identification technology, and particularly to an enhanced battery parameter identification method based on sinusoidal pulse current. The method comprises the following steps: S101. Building a test platform and collecting data; S102. Modeling the battery equivalent circuit; S103. Population initialization; S104. Generating sinusoidal excitation current; S105. Inferring response voltage; S106. Constructing the response voltage envelope; S107. Optimization objective based on the sum of squares error; S108. Constructing auxiliary optimization objective based on sinusoidal features; S109. Genetic evolution; S110. Termination condition judgment and optimal solution output extraction. This method fully and differentially excites the dynamic processes of different time constants within the battery. Without increasing any measurement hardware or testing costs, it improves information quality from the data source, opening a new path for achieving higher accuracy and greater robustness in parameter identification. This significantly improves the convergence efficiency and recognition accuracy of battery parameter identification.

The invention relates to a real-time three-dimensional modeling system fused with unmanned aerial vehicle target recognition, and relates to the technical field of unmanned aerial vehicle application, and the system comprises a data collection module which is used for obtaining multi-source sensor data; the real-time modeling module is used for constructing a real-time three-dimensional model of the environment The target identification module is used for identifying a specific target in the scene and associating the specific target with the three-dimensional model. The information quality evaluation module is used for quantifying the observation information quality of the recognized target in real time according to factors such as the shielding degree, the resolution ratio and the observation angle; and when the information quality evaluation result is lower than a preset threshold value, the flight path planning module is triggered to autonomously plan a dynamic flight path going to the optimal viewpoint, and the flight control module executes the flight path and drives the unmanned aerial vehicle to perform supplementary data acquisition. According to the method, the autonomous closed loop of the perception evaluation decision action is established, so that active and intelligent fine modeling of the key target is realized, and the integrity of model information and the operation efficiency are improved.

The invention relates to the technical field of natural language processing and information retrieval, and discloses a retrieval enhancement generation method and device based on self-feedback driving, and the method comprises the steps: determining a to-be-queried problem through a pre-obtained query request; retrieving in the knowledge base according to the to-be-queried question, and generating an answer of the round of retrieval; calculating an information quality index of the current round of retrieval according to the answer and the to-be-queried question, and if the current round of retrieval is not the first retrieval, calculating an information gain rate of the current round of retrieval according to an information quality index of the previous round of retrieval and the information quality index of the current round of retrieval; and if the information gain ratio does not meet the preset condition, returning to determine the to-be-queried question in combination with the pre-acquired query request to perform the next round of retrieval, and stopping iteration and outputting the final answer until the information gain ratio meets the preset condition. The problems that an existing retrieval enhancement generation method depends on a fixed retrieval threshold value and lacks dynamic judgment capacity are solved, and the information utilization efficiency and the answer accuracy are improved.

The application provides an electric vehicle instrument electronic guide book implementation method and device and electronic equipment, and relates to the field of data processing. The working procedure position and working content of a worker in the production work of an electric vehicle instrument are acquired, and the worker wears AR glasses; electronic guide content is determined from the electronic guide book according to the working procedure position and working content, and a first image is generated; the first image is displayed to the worker through the AR glasses; quality detection information of the electric vehicle instrument sent by a detection device is received, the quality detection information includes operated step information, operated step completion time and defective product conditions, and the quality detection information is used to indicate whether the quality of the electric vehicle instrument is qualified; the quality detection information is marked in the first image to generate a second image; and the second image is displayed to the worker through the AR glasses. Through the reasonable combination of the electronic guide book and product traceability, the beneficial implementation of the electronic guide book is facilitated.

The invention relates to the technical field of industrial operation safety monitoring and hazard source intelligent identification, and discloses an operation site hazard point identification method based on an attention mechanism neural network model. Sensor nodes capable of outputting multi-modal data are arranged along the line, and original data containing spatial positions and multi-dimensional observation values are collected at a plurality of discrete sampling moments; standardizing nodes, time and feature dimensions, and obtaining a unified embedded vector by using linear projection; self-attention is constructed based on node correlation in each mode, and weighted output is generated; the modal information quality is evaluated according to the output distribution dispersion at each moment, fusion factors are obtained and weighted fusion is carried out to form a multi-modal feature matrix; a risk score and a threshold value are obtained according to fusion feature statistical deviation, a modal contribution degree is calculated, and a risk label is generated; and finally, constructing a structured output vector, forming a dangerous point and non-dangerous point set, and realizing coverage enhancement, cross-modal balance, adaptive fusion and explainable output closed loop.

This invention discloses a new source evaluation method for internet technology information, belonging to the field of new source evaluation technology. To address the problems in existing technologies, such as users struggling to quickly find valuable information from a large volume of technology news, significant differences in information quality between different sources, and the lack of an effective source evaluation system, which prevents users from filtering information based on source reputation, this new source evaluation method for internet technology information effectively improves information acquisition and processing efficiency, ensures content quality and credibility, enhances readability and usability through sentiment analysis, topic classification, and relevance calculation, supports users in making informed decisions, promotes a virtuous cycle in the information ecosystem through multi-dimensional considerations in the evaluation system, and dynamically optimizes source evaluation through a user feedback mechanism, enabling personalized content recommendations, meeting user needs, and improving satisfaction.

According to the intelligent model reliability evaluation system based on the improved D-S evidence theory provided by the invention, the Wasserstein distance is introduced as a quantification tool for conflicts among evidences, and the measurement can accurately reflect essential differences among probability distributions of different evidence sources; the problem that the measurement is inaccurate when the distribution pattern difference of the traditional Euclidean distance or Jousseme distance is large is solved; meanwhile, Dane entropy is introduced to evaluate the information quality and uncertainty degree of the evidence, cognitive uncertainty and random uncertainty are effectively distinguished, and an information theory basis is provided for subsequent evidence weighting; namely, the Wasserstein distance and the Dane entropy are fused in order to quantify the distribution difference between the evidences and the uncertainty of the evidences more finely and improve the robustness, the accuracy and the interpretability of a reliability evaluation result, so that reliable evaluation support can be provided for landing application of a high-risk intelligent system, and the reliability evaluation method has the advantages of being high in reliability and high in reliability. The method is especially suitable for reliability capability evaluation of intelligent models in high-risk scenes such as complex systems and industrial control systems.

The invention relates to a method for outputting information in an output device having a display area, in particular of a vehicle, wherein the method comprises: Receiving information and associated metadata that indicate the quality of the respective information, wherein the metadata includes at least one piece of information about the source of the respective information; Determining a quality level for the respective information based on the associated metadata; Selecting or rejecting information based on the quality level; and outputting the selected information in the output device, wherein the output is at least partially adapted to the quality level of the information, characterized in that the method further comprises: Providing a means for a user, particularly via an HMI interface, to influence the evaluation of different information sources.

The present invention provides a method, apparatus, and device for optimizing information quality based on federated learning. The method comprises: obtaining a root cause analysis request sent by a service provider, the root cause analysis request carrying service quality data and service configuration data when the service quality is abnormal; inputting the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning; an optimization strategy is generated based on the root cause analysis result; and the service quality and / or network quality are optimized based on the optimization strategy. The solution of the present invention optimizes service quality and / or network quality with high security, high accuracy, high efficiency, and a complete process.

The invention discloses a scene parallax information generation method and device and electronic equipment, and relates to the field of image analysis, and the method comprises the steps: generating first type parallax information through a laser radar and a binocular camera; generating second-class parallax information through a deep stereo matching large model; third-class parallax information with the density larger than a preset threshold value is generated through the stereo matching system; generating a pseudo depth truth value by adopting a monocular depth estimation algorithm; performing preliminary fusion on the first to third types of parallax information to generate a preliminary fusion parallax; obtaining a scale value and an offset value based on the sparse parallax value and the pseudo depth truth value of the preliminary fusion parallax, and then adjusting the pseudo depth truth value to obtain fourth-class parallax information; and fusing the fourth type of parallax information with the first to third types of parallax information to obtain target parallax information. The problems of insufficient generalization ability, low parallax information quality and insufficient density in a complex real environment due to domain offset and point cloud sparsity of an existing stereo matching method are solved.