24 results about "Depth mapping" patented technology

The application discloses a power transmission line sag identification and early warning method and device combined with zoom ranging, relates to the technical field of power transmission line monitoring, and comprises the following steps: installing a binocular camera system in a target identification scene, adjusting variable focus lens parameters, calibrating the binocular camera system by using a checkerboard, and constructing a disparity-depth mapping relationship; collecting Q times of synchronous images by using the calibrated system, performing multi-point disparity analysis, obtaining disparity analysis results of a first suspension point, a second suspension point and a sag lowest point, and screening and determining target suspension point and sag lowest point analysis results; taking the results as indexes, searching the disparity-depth mapping relationship, and obtaining depth information; calculating target suspension point ranging results according to the depth information, calculating a sag value and judging whether the sag value exceeds a threshold value, and if not, triggering early warning, so as to solve the technical problem that a traditional ranging mode cannot feed back sag changes in real time, and thus cannot effectively early warn safety hazards caused by excessively large sag.

The present application relates to a method for predicting pre-drilling stratigraphic profile based on multi-algorithm interpolation of well logging curves, comprising: predicting the top / bottom and thickness of each layer of the target well based on stratified data; aligning the normalized depth within the layer; predicting multiple well logging curves by using multi-algorithm (at least IDW and Kriging) interpolation; correcting the proximity and mapping the depth based on the well deviation trajectory; outputting the multi-depth reference predicted curve and the stratigraphic position QC result. Further comprising lithology profile prediction: obtaining the classification probability by using stratified supervised classification, and fusing the spatial prior probability obtained based on the lithology of adjacent wells and distance weight to output the lithology interval of the target well and the profile comparison chart. This method comprehensively utilizes the well logging curves of adjacent drilled wells, stratigraphic position stratified data, well location and well deviation trajectory information, and under the unified depth reference, performs multi-algorithm horizontal interpolation prediction on multiple well logging curves of the target well, and further combines supervised learning and spatial prior to realize lithology profile prediction, and outputs the stratigraphic profile results which can be used for pre-drilling decision and well location deployment.

This invention presents a heuristic recovery mechanism for malicious shortcut file parsing and threat detection. It employs a resilience-oriented parsing engine, utilizes a dynamic byte ordering mechanism to adapt to underlying data, leverages a sliding window algorithm to accurately locate variable-length field terminators to extract hidden strings, and employs a recursive parsing strategy to deeply traverse undocumented nested structures and extension blocks. Parsing anomalies are treated as threat features and archived during the parsing process, and an iterative extraction strategy with adaptive parameters ensures full data acquisition. Based on the comprehensive parsing results, a three-dimensional threat model encompassing deception, evasion, and execution is constructed, achieving a deep mapping of malicious shortcut files from binary structure to logical behavior. This invention not only significantly improves the parsing success rate of variant samples but also effectively extracts fingerprint features from automatically generated samples, providing key technical support for modern shortcut-based attack and defense countermeasures.

Described herein are machine learning based systems, methods, and non-transitory computer readable media for determining three-dimensional (3D) information of structures of a patterned substrate. The 3D information can be determined using a neural network to convert a two-dimensional image to a 3D image. In a method, the neural network is trained by supplying varying parameters to simulated images of a sample to model a real manufactured IC structure. The trained network generates a depth map from a newly supplied image of the sample to predict defect locations. The depth map can be converted to a binary map to predict defect size and location, which can be used to guide inspection of the sample using an inspection tool.

This application discloses a multi-scale structure-aware optical field depth estimation method and system, relating to the technical field of 3D image processing. The method includes: constructing a candidate neighborhood micro-image set; constructing and scoring a consistency scoring function for each candidate view in the candidate neighborhood micro-image set, selecting the candidate views with the highest scores to obtain an adaptive neighborhood micro-image set; projecting reference pixels onto each neighboring micro-image in the adaptive neighborhood micro-image set, calculating the matching cost, and performing robust aggregation based on fuzzy difference weights; performing structure-aware propagation repair, reconstructing the matching window and repairing the cost propagation chain; iteratively updating the disparity plane, outputting a disparity map for each reference micro-image, fusing the disparity maps into a virtual depth map, and mapping the virtual depth to a measurable object distance to obtain the final depth map. This application can improve the accuracy and completeness of multi-focal-length optical field depth estimation.

This invention discloses a method and system for detecting and tracking aerial buoys from unmanned aerial vehicles (UAVs) based on multi-source sensor fusion. The method includes: constructing a unified spatiotemporal reference for lidar, RGB-D camera, and UAV pose information; performing real-time target detection on RGB images using a lightweight YOLO network to obtain buoy semantic information; generating buoy candidate boxes using the DBSCAN density clustering method based on spatial cropping and voxel downsampling of lidar point clouds; and constructing a dual-channel detection structure based on point cloud clustering and U-depth mapping using depth images. Scale and semantic prior constraints are introduced to cascade and fuse the multi-source detection results, and Kalman filtering is used to continuously estimate the target motion state to obtain buoy state information. This invention combines the long-range geometric stability of lidar, the near-range sensing advantages of RGB-D cameras, and the semantic discrimination capabilities of RGB images to achieve reliable target detection and tracking under complex backgrounds and high-speed relative motion conditions.

The application discloses a layered 3D picture rendering method based on CSS and relates to the field of front-end webpage development and graphics technology.The method comprises the following steps: receiving a page request containing a layered rendering mark, obtaining an original picture and layered parameters after analysis; preprocessing the original picture according to the layered parameters, segmenting the original picture into multiple picture slice layers based on depth mapping information; creating CSS elements for each slice layer and configuring independent CSS 3D transformation attributes, placing the slice layers in different three-dimensional space depths by setting different translateZ values and rotate attributes of each layer; and finally, combining the CSS elements in a document object model to form a layered 3D picture structure with a spatial depth feeling.The application further dynamically updates the CSS transformation attributes of each layer by responding to user interaction to generate a parallax motion effect based on the depth of the layers.The application fully utilizes the browser native CSS 3D hardware acceleration capability, does not need to rely on heavy JavaScript libraries, can realize efficient and light webpage 3D picture rendering, significantly reduces the development complexity and performance overhead, and simultaneously provides rich visual and interactive experiences.

The application relates to the technical field of road maintenance, and discloses a road surface pit slot intelligent detection method and system based on visual intelligence, the method comprising the following steps: acquiring multi-angle images and performing illumination compensation, and extracting pit slot potential edges; fusing multiple views to obtain a depth mapping diagram, combining the edges to complete three-dimensional modeling and calculating volume shape parameters, and generating a repair priority sequence; tracking the depth evolution and edge degradation of the repair priority sequence, and outputting a dynamic monitoring report and a joint risk verification result; updating a database according to the risk and optimizing sensor parameters, and forming a next cycle collection strategy. The method can solve the problem of low detection precision in the prior art.

The application discloses a loess drilling core image-based lithology and sedimentary facies identification and layering method, and aims to solve the problems that loess drilling cores have broken cores, missing sections, insufficient recovery rates, and that core box segmented shooting and splicing result in difficulties in automatic alignment of core images and well depth, accurate positioning of layering boundaries, and continuous output of the layering boundaries. The application obtains a core image set and well depth metadata, performs image preprocessing and splicing according to the core box sequence, and establishes an image position coordinate sequence. Image coding network and metadata coding network are used to extract features and generate an alignment feature sequence through cross-modal fusion of cross-attention Transformer. A depth mapping function that satisfies a monotonic increasing constraint is solved under the constraints of a starting well depth, a terminal well depth, a coring recovery rate and a depth marker point, so as to realize the mapping of the image position to the real well depth. Then, the image features are aligned to the well depth sampling points, and a lithology and sedimentary facies probability sequence is output. Finally, the final layering boundary is determined under the constraints of a minimum layer thickness and adjacent layer section category transition, so as to realize the continuity constraint output of the layering result and improve the alignment accuracy and layering reliability.

The application relates to the technical field of computer vision, in particular to a three-dimensional positioning method and system of array elements, wherein the method comprises the following steps: acquiring a multi-view high-bit-depth gray image frame sequence of an array element to be measured; completing dynamic range conversion from high-bit-depth to low-bit-depth through adaptive bit-depth mapping enhancement; performing primary detection on the low-bit-depth image to obtain a candidate coded visual marker and a corresponding ROI region; performing local mapping enhancement on each ROI region to obtain an enhanced ROI image; performing secondary precision detection on the enhanced ROI image to output a target coded visual marker containing a target corner point and a single-frame pose; determining a corner point observation covariance according to the ROI dynamic range; constructing an incremental factor graph taking the single-frame pose as a state quantity and taking a re-projection error as an observation constraint, establishing an adaptive noise-weighted multi-frame joint optimization target; and performing optimization solving based on the optimization target to obtain three-dimensional coordinates of the array element in a world coordinate system. Through the application, the precision, consistency and real-time performance of three-dimensional positioning of the array element are improved.

This invention relates to the field of data processing technology, providing a rendering method, apparatus, device, and medium based on a panoramic resource service engine. It can reconstruct the spatial coordinate system of a UAV panoramic dome image based on surveying vector data, solving the visual displacement problem between the virtual redline and real-world features during 360° rotation. It performs attitude compensation rendering on the reconstructed dome image based on real-time viewpoint data and a nonlinear mapping model, ensuring the redline always conforms to the real-world terrain without floating. It uses pixel-by-pixel depth mapping to perform occlusion culling on the dynamic redline layer, and then composites the occluded dynamic redline layer with the reconstructed dome image, enabling the redline to be deeply embedded in the real-world scene, significantly enhancing visual realism. Finally, it performs interactive rendering of the target sub-region based on resource flow information and a nonlinear mapping model, thereby achieving an interactive panoramic display of resource flow information.

The application discloses a tunneling target positioning method based on complex loose body multi-modal data survey, relates to the technical field of underground space artificial intelligence survey, and comprises the following steps: synchronously collecting multi-modal sensing data and tunneling working condition data in a tunneling process and performing adaptive noise reduction processing; extracting an energy envelope feature sequence of the multi-modal sensing data and performing depth mapping alignment; inputting the aligned multi-modal data features into a double-branch feature coding network, introducing a semantic prior vector as a context constraint, distributing fusion weights of the multi-modal data features through a cross attention mechanism, and outputting a high-dimensional fusion feature tensor; and constructing a physical information constraint neural network, taking electromagnetic and acoustic physical equations and auxiliary sensor data as constraint terms of a loss function for calculation, and outputting three-dimensional positioning coordinates of a tunneling target in a tunnel global coordinate system. The method eliminates time-depth conversion errors caused by complex loose body multi-variation.

This invention discloses a naked-eye 3D augmented reality interactive display system, relating to the field of naked-eye 3D display and real-world interaction technology. The invention acquires and fuses user limb depth images through a multi-source depth data acquisition module, completes the mapping of depth and parallax and viewpoint correction through a parallax physical depth mapping module, divides regions to determine occlusion types through an occlusion relationship determination module, generates a dynamically smooth occlusion layer through an adaptive occlusion layer generation module, and displays the occlusion effect fusion display module by superimposing the occlusion layer with multi-viewpoint virtual images. This effectively solves the problems of inaccurate occlusion relationships between user limbs and virtual objects and visual abruptness in naked-eye 3D interaction, significantly improving the spatial consistency and visual smoothness of the interaction, and achieving a realistic and natural naked-eye 3D augmented reality interactive experience.

This invention discloses a method, apparatus, and electronic device for reconstructing the three-dimensional morphology of an electrode material surface. The method includes: determining the total depth confidence index corresponding to multiple candidate depth values ​​in the (t-1)th iteration; determining the depth probability value of a corresponding pixel in the tth iteration for each of the candidate depth values ​​based on the total depth confidence index and pixel sharpness; determining the iterative depth value of the corresponding pixel based on the corresponding depth probability value, until a target condition is met, to obtain the target depth value; combining the target depth values ​​corresponding to multiple pixels to obtain a depth map; and reconstructing the three-dimensional morphology of the electrode material surface based on the depth map. This invention solves the technical problem in related technologies where inaccurate determination of the depth value of the electrode material surface leads to low accuracy in the three-dimensional morphology reconstruction of the electrode material surface.

The application discloses an error online detection and correction system and method for a mechanical accessory assembly process, relates to the field of automatic assembly of heavy load hydraulic cylinders, and aims at the problem of visual profile distortion caused by eccentric oil film in high-reflective plunger rod blind insertion. The system calculates three-dimensional coordinate drift distance by combining a boundary splitting module with a depth constraint and a mapping matrix, splits out a real rod profile exposure point set and a meniscus edge covering point set, a direction inversion module reduces the dimension of the point set, constructs a first contact inversion direction vector according to the spatial opposite direction of the two centers, a directional correction module translates along the opposite direction of the vector and slightly tilts the posture based on the extracted net exposure amount and topological inclusion relationship, and a blind insertion execution module uses absolute depth mapping to guide space oblique interpolation under the condition of frozen visual feedback. The application excludes optical outward expansion illusion from the bottom algorithm, converts visual features into anti-collision vectors, and realizes high-precision soft assembly in a deep limited blind area.

The present application relates to the technical field of image acquisition, in particular to a rigid endoscope image acquisition method and system based on eyepiece imaging, comprising: receiving a first light path image and a second light path image transmitted through an objective lens by an eyepiece assembly of a rigid endoscope, and performing photoelectric conversion on the first light path image and the second light path image to obtain corresponding first original electric signals and second original electric signals; inputting the first original electric signals and the second original electric signals into a pre-trained image denoising model for denoising processing to obtain a first denoised image and a second denoised image. The present application generates a real-time depth mapping graph by combining posture sensor data, so that the endoscope can accurately reflect the distance distribution between the front end and the photographed tissue when observing the tissue, which helps doctors obtain more accurate spatial information.

This application provides a flight verification system and method for an unmanned aerial vehicle (UAV) emergency landing assistance system, applied in the field of artificial intelligence technology. It includes: a UAV model building module constructing a six-degree-of-freedom (DOF) dynamic model of the UAV; a ground simulation verification module simulating the six-DOF dynamic model to obtain the UAV's real-time attitude vectors and simultaneously rendering matching initial visual images and depth maps; then, combining the UAV's external disturbance states to construct the UAV's global state space in a Markov decision model, followed by strategy training; and finally, through fault injection simulation optimization of the decision sequence, outputting a first optimal avoidance flight path; when the consistency score between the first optimal avoidance flight path and the second optimal avoidance flight path output by the emergency landing assistance system exceeds a preset score threshold, the flight verification result of the emergency landing assistance system is determined to be qualified. Under the premise of ensuring safety, the flight verification of this emergency landing assistance system is achieved scientifically and objectively.

This invention provides a method, apparatus, equipment, and storage medium for assessing traffic resilience and delay costs under rainfall-induced urban flooding. The method includes: acquiring rainfall data for a target area; performing runoff generation and confluence simulations based on the rainfall data to obtain the water depth; determining a safe driving speed corresponding to the water depth based on the water depth and a speed-depth mapping model; the speed-depth mapping model determining a safe driving speed while ensuring vehicle handling stability, wherein the safe driving speed is obtained based on the road water depth and a scaling factor; obtaining a travel demand matrix based on the spatial distribution of the population in the target area; determining traffic delay time based on the safe driving speed and the travel demand matrix; and obtaining traffic delay costs based on the traffic delay time. This invention can improve the efficiency and accuracy of assessing traffic resilience and delay costs under rainfall-induced urban flooding.

The malicious shortcut file analysis and threat detection method of heuristic recovery mechanism, the analysis engine guided by resilience, adopts the dynamic byte sequence processing mechanism to adapt to the underlying data, uses the sliding window algorithm to accurately locate the variable length field terminator to extract the hidden string, and through the recursive analysis strategy, deeply traverses the undocumented nested structure and extension block. In the analysis process, the analysis exception is regarded as a threat feature for archiving, and the parameter adaptive iterative extraction strategy is adopted to ensure the full acquisition of data. Based on the complete analysis result, a three-dimensional threat model covering deception, evasion and execution is constructed, and the deep mapping of the malicious shortcut file from the binary structure to the logical behavior is realized. The present application not only significantly improves the analysis success rate of variant samples, but also effectively extracts the fingerprint characteristics of automatically generated samples, providing key technical support for modern shortcut-based defense confrontation.

This invention relates to the field of intelligent construction control technology, specifically to a rapid construction method for collaborative initial support of tunnel equipment based on artificial intelligence. The method includes the following steps: acquiring parameter ratios to construct a coupled stiffness gradient; constructing a basic reward matrix based on a Markov decision process and calculating a negative penalty scalar; quantifying collision avoidance constraint factors using Euclidean distance; generating a collaboratively updated reward matrix and filtering and outputting target travel increment commands; and executing underlying bus conflict detection to establish a collaborative operation sequence. In this invention, a gradient system constructed by calculating multi-source physical parameters overcomes the lag in experience-dependent perception, deeply maps and predicts the deformation trend of surrounding rock and transient pressure changes, quantifies interference risks through spatial coordinate calculations, deeply integrates collision avoidance factors with the control sequence, and executes underlying time-domain translation scheduling to achieve interlocking of actions and obstacle avoidance coordination among multiple devices in confined working environments, completely eliminating the risk of trajectory collisions in parallel operations.

The application discloses a target point driven arbitrary cross-section CT image visualization method and system, and the method comprises the following steps: receiving a selected point coordinate, using the speed and acceleration of a historical trajectory to predict a depth mapping to obtain an initial target point; taking the initial target point as a starting mass point, calculating a virtual gravity resultant force based on a gray histogram multi-peak detection and moving the mass point, determining an anchoring target point through oscillation suppression; taking the anchoring target point as a cross-section origin, generating a cross-section according to a gradient direction histogram adaptive fusion user normal vector; constructing a virtual spring system based on a projection point offset to correct a rotation angular velocity, updating the cross-section posture; adopting a neighborhood prediction and segmented variable step length strategy to advance a sampling ray along the normal vector, and stopping when the accumulated gray difference reaches a threshold value; finally, generating a cross-section image through median filtering, adaptive thickness Gaussian weighting and contrast stretching resampling. Precise positioning, adaptive cross-section orientation, efficient sampling and high-quality imaging are realized, and the interactive efficiency is significantly improved.

This invention discloses a method for determining the pass / failability of a tensile force-stretch ratio curve based on image recognition, comprising the following steps: S1, collecting tensile force-time series data under multiple sets of preset process parameters; S2, extracting the stretching segment data, converting the tensile force-time data into tensile force-stretch ratio data, and plotting a curve based on the time series data; S3, constructing a curve classification model based on a dual-path feature extraction network; S4, iteratively training the curve classification model, and using the trained curve classification model for determining the pass / failability of the tensile force-stretch ratio curve. This invention utilizes the tensile force-stretch ratio curve generated during the stretching process of a biaxial stretching apparatus to establish a deep mapping from the production dynamic curve to the final quality index of the film, thereby determining the process pass / failability.