446results about "Picture interpretation" patented technology

The invention discloses a multi-agent collaborative visual navigation reasoning enhancement method and system, and the method comprises the steps: an inference agent analyzes input data, and generates structured output containing a reasoning process and an initial answer; independently analyzing the input data by the evaluator agent, generating an evaluation answer and reviewing the output of the inference agent; comparing whether the initial answer is consistent with the evaluation answer, and if yes, outputting a final answer; if not, calling a bifurcation feedback agent, and generating a core bifurcation point abstract and a visual evidence verification plan; and calling a final decision maker agent to directionally retrieve the video evidence according to the visual evidence verification plan, and outputting a final answer. Four roles of an inference person, an evaluator, a divergence feedback person and a final decision maker are introduced, a complete'generation-evaluation-feedback-optimization 'decision link is constructed, and the cognition and inference accuracy in a complex scene is remarkably improved.

The invention relates to the technical field of geographic information engineering, in particular to a surveying and mapping geographic system based on a remote sensing technology, which is characterized in that a multi-modal remote sensing data acquisition module synchronously acquires satellite remote sensing images, aviation LiDAR point cloud, unmanned aerial vehicle oblique photography and ground Internet of Things sensing data; the data fusion collaboration module carries out geographic coordinate calibration and position coding, and generates a surveying and mapping data set with unified precision in combination with an improved Mamba fusion model; the dynamic terrain high-precision inversion module performs three-dimensional terrain reconstruction, and completes terrain deformation monitoring in combination with an InSAR interference measurement technology and a time sequence analysis algorithm; the real-time error correction module constructs a BP neural network optimized by an ISSA algorithm, and dynamically corrects a terrain three-dimensional model and a terrain deformation result; the intelligent visual application module constructs a digital twinborn simulation model and automatically generates a standardized surveying and mapping result report. Therefore, the problems of weak multi-source data fusion, low terrain inversion precision and the like in the prior art are solved.

The invention provides a method and a device for installing, positioning and monitoring a laminated slab in a beam. Relates to the technical field of building construction. According to the specific implementation scheme, the method comprises the steps that reverse modeling is conducted on a prefabricated laminated slab component through laser scanners placed at multiple target positions, and a point cloud model is generated; inputting preset data of the prefabricated laminated slab component into the building information model to generate a target preset model; and the point cloud model is matched with a target preset model, and when it is detected that the matching degree of the point cloud model and the target preset model is smaller than a first preset threshold value, multiple installation point positions for beam-entering installation of the prefabricated laminated slab component and normal state parameters of the multiple installation point positions are determined based on the point cloud model and the target preset model. According to the technical scheme, the accuracy and construction efficiency of installation and positioning of the laminated slab into the beam can be improved, structural hidden dangers caused by installation dislocation are avoided, it is guaranteed that the laminated slab and the beam form a reliable overall stress system, and the stability and safety of a building structure are enhanced.

The invention discloses an unmanned aerial vehicle multispectral system error calibration and correction method based on a Boolean model, and belongs to the technical field of photogrammetry and remote sensing. Comprising the following steps: acquiring a multispectral image, acquiring original POS data, and performing band registration preprocessing on the image to generate a first orthoimage to confirm a system error magnitude; carrying out bundle method strict adjustment by using ground control points to obtain reference POS data, and solving seven-parameter conversion parameters; correcting the original POS data by using seven parameters and generating a high-precision orthoimage; and generating a simplified multispectral camera calibration and verification formula according to the difference value of the POS data before and after correction. Complex coordinate conversion is simplified into fixed value addition and subtraction operation, one-button calibration and batch correction of system errors of the multi-spectral camera of the unmanned aerial vehicle are achieved, the positioning precision is improved to the centimeter level from the meter level, dependence on ground control points is greatly reduced, and an efficient standardized solution is provided for large-scale engineering data processing.

This invention relates to a system and method for non-invasive detection of gravesites and archaeological features using multimodal remote sensing and machine learning. Remotely sensed datasets, including RGB, multispectral, hyperspectral, LiDAR, and thermal imagery, are orthorectified, mosaicked, and subdivided into tiled image segments. Features are labeled through manual annotation of visible markers and environmental signatures and expanded via iterative augmentation. A supervised pipeline trains computer vision models, such as YOLO-based detectors, in parallel with tabular models derived from spectral indices (NDVI, NDRE), LiDAR elevation derivatives, and thermal anomalies. Inference outputs are cross-validated against thresholded evidence layers to reject false positives and upgraded when spectral, spatial, and thermal evidence align. Validated detections are exported as GIS-compatible layers with confidence scores and metadata. The system provides a scalable, replicable tool supporting archaeologists, Indigenous communities, and planners in cemetery investigations, cultural resource management, and humanitarian searches for unmarked or clandestine graves.

The invention provides a data processing method and device based on a smart city, and relates to the technical field of data processing.The method comprises the steps that a city-level spatial data base is constructed through oblique photography of an unmanned aerial vehicle, and a unified and real spatial foundation is formed by generating a live-action three-dimensional model, a digital orthoimage and a digital elevation model; on this basis, a building base and a building white model are extracted, spatial logic verification is carried out on multi-source business basic data, a standard address system stably associated with a building three-dimensional entity is further constructed, precise spatial anchoring of governance objects such as population, houses and units is achieved, and the method is suitable for mass production. And finally, the associated data is uniformly converged to a city information model platform, and reliable and updatable spatial data support is provided for smart city application. According to the invention, the timeliness of smart city data processing in a dynamic earth surface deformation scene can be improved.

The invention provides a cable-stayed bridge cross brace hoisting monitoring and early warning method, and the method comprises the steps: building a finite element model based on the design parameters of a cable-stayed bridge and a cross brace, and calculating the safety envelope of the displacement, rotation angle and stress of a control point; displacement, strain and cable force sensors and industrial cameras are arranged in a cross brace and an anchoring area, and data and images are collected. Inputting the image into a first convolutional neural network model, extracting attitude parameters such as vertical and lateral offsets and torsion angles of each control point, and adapting to different imaging scales through a scale adaptive convolution module; and aligning and splicing the attitude parameters with sensor displacement, strain and cable force indexes according to time to form a feature sequence, inputting the feature sequence and safety envelope parameters into a second convolutional neural network risk assessment model, outputting a comprehensive risk index reflecting deviation from safety envelope, and generating construction treatment instructions such as hoisting speed reduction and hoisting point stress adjustment. According to the monitoring and early warning method, the construction safety is greatly improved, and monitoring and early warning of the whole process of cross brace hoisting are achieved.

The invention discloses a visual imaging method and system combined with unmanned aerial vehicle multi-temporal automatic comparison, and the system comprises an unmanned aerial vehicle aerial survey module which is used for carrying out the repeated aerial survey of the same target region at different time points, and obtaining multi-period image data; the data transmission and storage module is used for transmitting the multi-period image data acquired by the unmanned aerial vehicle aerial survey module in real time; the method has the beneficial effects that the aerial survey of the unmanned aerial vehicle supports dynamic route adjustment, multi-sensor cooperative acquisition is realized, the overlap ratio of multiple aerial survey paths is high, the image data integrity and consistency are high, and the manual intervention and data supplementary acquisition cost is greatly reduced; multi-stage image automatic registration, feature extraction and change detection are realized through deep learning, the processing precision is high, a large amount of manual operation is not needed, and the efficiency is remarkably improved; three-dimensional modeling and multi-mode comparison are supported, dynamic changes of ground objects can be clearly displayed, interaction functions are rich, and the requirements for single-person detail checking and multi-person collaborative rechecking can be met.

The invention relates to a target tracking optimization method and system based on image recognition laser ranging. According to the method, when the intensity of a signal returned by laser ranging operation is lower than a preset threshold value, a candidate target is determined as a non-target and a non-target mark is added, and in subsequent image processing, when a candidate target associated with the non-target mark is identified again, the identification confidence of the candidate target is reduced, so that the identification confidence is lower than a threshold value for triggering ranging. According to the method and the system, repeated laser ranging operation on a non-target object can be avoided, so that unnecessary computing resource consumption is reduced, the efficiency and the accuracy of the system for tracking a real target are improved, particularly, tracking interruption can be effectively avoided when the target moves highly dynamically, and the robustness and the reliability of the system are improved.

The invention relates to the technical field of intelligent inspection of highway facilities, in particular to an unmanned aerial vehicle aerial photography highway separation gate damage identification method, which comprises the following steps of: firstly, generating an imitated flight route matched with linear distribution of a separation gate based on GIS (Geographic Information System) geographic information and a digital elevation model of the highway separation gate; then the unmanned aerial vehicle is controlled to autonomously fly along a route, a high-resolution sequence image containing an isolation fence stand column and a mesh is collected at a preset equidistant shooting point, and positioning and attitude determination data are synchronously recorded; a pre-training deep learning model is used for detecting the inclination angle of the stand column and the mesh damage area, and abnormity is marked; and finally, fusing the positioning and attitude determination data and the identification result, calculating the three-dimensional geographic coordinates of the abnormal points through a space forward intersection algorithm, and generating an isolation gate anomaly report with the geographic coordinates. The problems of low efficiency and poor precision of traditional inspection are solved, the abnormity identification and positioning precision is improved, and a precise basis is provided for isolation gate maintenance.

The invention discloses a space-based double-satellite asynchronous detection multi-target correlation positioning system and method, the system comprises space-based optical satellites and a central processing unit, each space-based optical satellite is equipped with a large-view-field optical electro-telescope, and the large-view-field optical electro-telescope is used for carrying out image acquisition on a space target; and the central processing unit comprises a computer, the computer is in communication connection with the space-based optical satellite, and an image processing module, an angle measurement data generation module, a data association module and a target positioning module are integrated in the computer. According to the characteristics of space-based double-star asynchronous observation, a complete process from image processing, angle measurement data generation, time series data association to final positioning is designed, and the method is particularly suitable for monitoring and positioning a large number of space non-cooperative moving targets; by introducing time registration and sequence geometric consistency judgment, the association problem caused by asynchronous double-star observation time is effectively solved, and the accuracy and robustness of data association in a multi-target scene are improved.

The invention provides an underwater terrain inversion method based on a height measurement satellite, belongs to the technical field of remote sensing geoscience application, and constructs an automatic processing flow integrating space screening, elevation correction, signal denoising, double-strategy surface and bottom recognition and physical correction. The method is suitable for underwater topography generation in complex water body environments such as lakes and reservoirs, and the efficiency of underwater topography data acquisition is greatly improved. A strategy of combining adaptive kernel density estimation and clustering analysis is adopted to remarkably improve the water surface and water bottom signal recognition capability, and a multi-temporal collaborative analysis and refractive index physical correction model is introduced to realize high-precision underwater terrain extraction of water bodies such as inland reservoirs. The method has obvious advantages in the aspects of automation degree, application range and result stability, and has important engineering application and popularization value.

The invention relates to the technical field of geological disaster monitoring, and discloses an unmanned aerial vehicle aerial survey intelligent identification and grading method and system for geological disaster hidden danger points, and the method comprises the steps: converting a pixel-level identification result into hierarchical data which retains a spatial topological structure through constructing a confidence quad-tree; establishing a weighted decision grid based on aerial survey parameters to realize coupling modeling of observation conditions and confidence uncertainty, calculating topology difference to generate a spatial difference field to quantify the deviation degree of a confidence field and a standard grading decision surface, and finally executing self-adaptive grading judgment to output a risk grade result with a confidence label; according to the method, the problem of confidence field and decision boundary space topology mismatch caused by truncation of confidence continuous measurement information in a grading stage in a traditional serial architecture is solved, systematic underestimation of misjudgment risks in observation condition limited areas such as alpine and gorge areas is avoided, and consistency of actual credibility of a risk quantification result and a nominal grade is realized.

The invention relates to the technical field of photogrammetry, and discloses a photogrammetry space reconstruction method based on multi-source image fusion, which comprises the following steps: extracting an interior orientation element and an initial exterior orientation element, and establishing an image beam intersection constraint intensity model in an object space to calculate a constraint weight; monitoring a weight space change rate to identify a heterogeneous image connection area and extracting a high-weight image bundle as a geometric anchor point; a differential regularization item is applied to geometric anchor points to construct a stable geometric skeleton, a weight damping factor is utilized to restrain a low-weight image beam to execute conformal mapping correction, the image beam constraint strength is represented, the condition number of an adjustment method equation matrix is improved, numerical oscillation caused by difference of different-source sensors is restrained, geometric step of a fusion interface is eliminated, and the conformal mapping correction accuracy is improved. And geometric closing and seamless stitching of a multi-source image under a unified framework are realized.

The invention relates to a three-dimensional map surveying and mapping method, device and equipment based on an unmanned aerial vehicle and a medium. The method comprises the following steps: acquiring a global remote sensing image of a to-be-mapped area, and performing scene semantic analysis on the global remote sensing image to obtain scene semantic information; the global remote sensing image is collected by a remote sensing satellite; generating an initial flight path according to the scene semantic information, and obtaining three-dimensional surveying and mapping data collected by the unmanned aerial vehicle along the initial flight path; the three-dimensional surveying and mapping data comprises multi-view image data and corresponding positioning and attitude determination data; generating an initial point cloud of the to-be-mapped area based on the three-dimensional mapping data; carrying out ground feature category division on the initial point cloud to generate a hierarchical semantic point cloud; and generating a three-dimensional map model corresponding to the to-be-mapped area based on the hierarchical semantic point cloud. By adopting the method, the layered three-dimensional map with surface feature semantic information can be generated, and dynamic optimization can be realized.

The invention discloses a substation operation scheme compilation method based on a three-dimensional model, and relates to the technical field of power operation maintenance, and the method comprises the steps: collecting multi-source data, constructing a three-dimensional digital twinborn body, accessing real-time data, and forming a digital twinborn platform; querying an overhaul task and performing text structured processing to form an overhaul equipment set; carrying out special analysis and providing a special analysis result; the maintenance scene is displayed in the digital twin platform, and a maintenance operation scheme is generated and output; and after the maintenance operation is finished, field maintenance data are returned, and the digital twin platform data are updated. According to the method, the three-dimensional digital twinborn body which is synchronously mapped with the transformer substation and dynamically updated is constructed, so that a maintenance operation scheme is compiled based on a real-time field state, the problem that a traditional three-dimensional visualization system is subjected to static display and is disjointed with the field is solved, traditional tedious and error-prone steps depending on artificial experience are converted into automatic algorithm calculation, and the maintenance efficiency is improved. The time required for generating special analysis is greatly reduced, and the operation and maintenance cost is reduced.

The invention discloses an unmanned aerial vehicle adaptive three-dimensional reconstruction method based on semantic and geometry collaborative perception, and the method comprises the steps: obtaining image data collected by a binocular camera in real time, executing semantic segmentation in parallel to generate a real-time semantic graph, and calculating a geometric complexity graph based on a depth graph; according to a preset association rule base, fusing the real-time semantic graph and the geometric complexity graph, and identifying a region of interest; dynamically generating an optimized flight path on the basic route based on the position and the attribute of the region of interest; and controlling the unmanned aerial vehicle to fly according to the optimized flight path, adjusting sensor parameters to execute high-density data acquisition, and transmitting the acquired data and metadata to a ground station. According to the method, adaptive optimization of three-dimensional reconstruction of the unmanned aerial vehicle is realized by integrating semantic segmentation, geometric analysis and dynamic path planning.

The invention relates to a method of encoding image data representing a geographical line, GL, or its surroundings, and a data processing system being configured to perform the encoding. The method comprises applying a data compression process to reduce the spatial resolution of a given image represented by the image data. Specifically, the image may form an imagery layer of an imagery stack comprising various imagery layers, each imagery layer representing an individual image, e.g. of a same geographical region. The data compression process comprises: generating two or more sublayers of the image by resampling at least one region of the image at a respective lower spatial resolution than its original spatial resolution, while maintaining or resampling the spatial resolution of at least one other region of the image such that, when the sublayers are overlapped to jointly define a combined multi-resolution representation of the image, the average resolution of the combined multi-resolution representation is lower than the original resolution of the image.

The invention discloses a geological disaster hidden danger point rapid identification method based on aerial survey of an unmanned aerial vehicle, and relates to hidden danger point rapid identification. An airborne camera image sequence, the acceleration and angular velocity of an inertial measurement unit, the propeller rotating speed, the flight height, the wind speed and direction and a machine body vibration signal are collected, and image exposure time is used as a unified time axis; the method comprises the following steps: forming a time sequence sample, establishing rotor underwash near-earth boundary layer momentum entrainment prior based on the time sequence sample, calculating underwash dynamic pressure field time sequence estimation, a particle mobilization probability time sequence and an equivalent micro-attitude disturbance spectrum, performing sub-pixel level optical flow and phase accumulation analysis on an image sequence, performing time cross-correlation with the equivalent micro-attitude disturbance spectrum, and obtaining a target object. Coupling evidence is obtained, joint judgment of the triggering intensity and the coupling degree is carried out, a flight parameter self-adaptive fine tuning strategy is triggered, space risk analysis is carried out on the image after compensation reconstruction, a space risk calibration graph is generated, and geological disaster hidden danger points in a critical stable state can be effectively recognized.

Identifying hazard trees is described. An example method includes receiving multiple images for a geographic area that includes multiple electrical assets of a power distribution infrastructure. Pixels of the multiple images are classified as hazard tree pixels or as non-hazard tree pixels using one or more convolutional neural networks. Multiple polygons are generated based on the hazard tree pixels, a polygon corresponding to one or more hazard trees in the geographic area. A height of a polygon and a distance from the polygon to an electrical asset is determined. Based on the height and the distance, the one or more hazard trees corresponding polygon are determined to represent a potential hazard to the electrical asset. A notification of the potential hazard to the electrical asset is generated and provided.

The invention relates to the technical field of image processing, in particular to a mine safety monitoring method and system for an open-pit mining area. The method comprises the following steps: firstly, constructing different sliding windows to slide on a wedge-shaped slope image, extracting periodic characteristics of a slope surface by using similarity evaluation indexes, and comprehensively evaluating the similarity of each region of the image; secondly, determining periods of the image in vertical and horizontal directions based on similarity evaluation indexes, and optimizing block grid parameters according to the periods, so that the algorithm processing scale is matched with the periodic characteristics of the image, and the pertinence and detail representation of image enhancement are improved; and finally, the crack risk evaluation index is calculated by using the enhanced image, so that the accuracy and reliability of crack risk evaluation are remarkably improved, and the precision and robustness of slope landslide disaster monitoring are enhanced.

A system and method for achieving a high confidence threshold when asserting that the runway is clear of obstacles. A landing aircraft periodically scans runway regions to confirm, track, and propagate trajectories of self-reporting and non-cooperative objects. An incurring object, appearing as a very tiny artifact, is unlikely to be detected in every sensor frame. The system and method propagate the trajectories of the incurring objects to confirm that the objects will clear the runway before the aircraft lands by tracking the objects and using a probabilistic decision maker.

An image processing device 20 includes a communication unit 21 and a controller 22 (processor). The communication unit 21 acquires a standard image and a reference image captured by a stereo camera 10. The controller 22 calculates a disparity based on the standard image and the reference image. The controller 22 calculates, at each pixel position, a cost value representing a degree of difference for each pixel position between the standard image and each of multiple images obtained by displacing the reference image by multiple different deviation amounts in a direction corresponding to a baseline length direction of the stereo camera 10. The processor aggregates the cost values using dynamic programming and calculates a disparity for each pixel position based on the aggregated cost values for each different deviation amount. The multiple different deviation amounts are discrete values at uneven intervals.

This application relates to the field of image measurement technology and proposes an automatic detection, tracking, and positioning method for aerial targets, applicable to high-precision perception and stable locking of low-altitude and near-altitude moving targets in complex environments. This invention employs a system redundancy and fusion detection approach, utilizing both a visible light camera and a lidar sensor to construct two sets of detection sensors with different modes for target detection. The data from these different modes are then fused, enabling automatic detection, tracking, and positioning of small aerial targets at distances greater than 500 meters in the far field. This provides target positioning information for subsequent precise detection, focusing, and real-time tracking display.

The present application relates to the technical field of automation control, and especially relates to a box workpiece pose measurement method based on point features, comprising calibrating a monocular camera by Zhang Zhengyou calibration method, obtaining internal parameters and distortion parameters of the monocular camera; selecting four corner points of four box workpieces in a three-dimensional space as feature points in an image, establishing a world coordinate system, a camera coordinate system, an image coordinate system and a pixel coordinate system, and obtaining three-dimensional coordinates of the feature points in the world coordinate system; using the point features to perform image processing on the preprocessed image, obtaining two-dimensional coordinates of the feature points in the pixel coordinate system; combining the three-dimensional coordinates of the feature points in the world coordinate system, the two-dimensional coordinates in the pixel coordinate system and the internal parameters of the monocular camera, and using a PNP measurement method to solve the pose information of the box workpiece. The present application provides a solution to the problem that the existing pose measurement method is complicated and cannot measure the pose of small box workpieces.

A measurement system has an extraction section and an alignment section. The extraction section extracts a point group of a first marker from a measurement point group including a point group of a measurement object and a point group of the first marker arranged at a known position with respect to the measurement object. The alignment section aligns the point group of the measurement object with a known point group by alignment of the point group of the first marker with a point group of a second marker, the position of which is known with respect to the known point group of the measurement object.

A device and method for identification of at least one object in an image registered with an image registration device, wherein each of the at least one object is provided with a wireless tag, the device comprising at least one sensor for registering a wireless signal from the wireless tag and for registering at least one direction to the wireless tag. The wireless signal comprises information for identification of the object.

The invention discloses a multi-modal visual language navigation method based on a dynamic environment knowledge graph and related equipment, and can be applied to the technical field of artificial intelligence. After the real-time environment image of the area to which the target robot belongs is captured through the single camera, feature extraction is carried out on the real-time environment image to obtain the image features, the to-be-processed category corresponding to the target object is coded to obtain the semantic code, and after the spatial features of the target object are extracted, the target robot is obtained. Constructing a dynamic environment knowledge graph according to the confidence, semantic coding and spatial features of the target object in combination with an external knowledge base, extracting high-dimensional graph features of the dynamic environment knowledge graph, and performing implicit fusion and display modeling in combination with image features and language features corresponding to the target instruction to obtain a navigation state vector containing long-term memory; therefore, the moving operation of the target robot can be controlled based on the navigation state vector, efficient and explainable navigation reasoning is further realized, and the language navigation accuracy is improved at relatively low cost.

A system and method for monitoring the shape and size (deformation) of a dyke (7) or similar terrain is disclosed. This method employs the monitoring of positions of settlement plates (12) by an unmanned aerial vehicle UAV (5) in combination with ground based GNSS sensors (11) and barcode markers (3, 4).