417results about "Details involving processing steps" patented technology

The invention discloses an AR home experience method in a large scene. On the basis of combination of a natural feature identification-based three-dimensional registration method and a binocular tracking positioning and local mapping method, the camera attitude is estimated by using feature points of a real-time scene and corresponding three-dimensional points thereof under the binocular trackingpositioning and local map construction technology. According to the mode, on-site environment features shot in real time are used as recognition tracking objects, a virtual home model can still be normally positioned and tracked under the condition that no identification graph exists, the problems that an existing AR home experience application is small in use range and poor in stability are solved, and therefore the AR home experience of virtual and real fusion can be met in a wider range and more truly.

The invention provides a point cloud downsampling method applied to unmanned aerial vehicle obstacle avoidance, and aims to solve the problems of low mapping efficiency and geometric detail loss caused by large point cloud data volume and non-uniform density in environmental perception. The method comprises the following steps: preprocessing an original point cloud through a regular grid to balance density, and extracting an effective flight area of the unmanned aerial vehicle by adopting bidirectional cloth simulation filtering and combining with a flight height; multi-level geometric features including elevation gradient, curvature and roughness are calculated, and point cloud priority scores are fused and generated based on principal component analysis; voxel grids are divided according to the size of the unmanned aerial vehicle, the sampling rate of each voxel is adaptively set in combination with the priority, farthest point sampling is executed in the feature space to select representative points, and finally a simplified point cloud model is output. The method can be integrated to a mapping module of a multi-aircraft cooperative obstacle avoidance and planning system, is suitable for unmanned aerial vehicle path planning tasks in a complex three-dimensional environment, and has the characteristics of light weight, accuracy, high efficiency and the like.

The invention relates to a myocardial resection area determination method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring a heart scanning image, and determining a heart three-dimensional model according to the heart scanning image; determining a heart short-axis section and a heart long-axis section according to the heart three-dimensional model; the clock direction angle of the scalpel is determined on the short axis section of the heart; based on the clock direction angle, the excision depth of the scalpel is determined on the long-axis section of the heart; and performing myocardial resection on the heart three-dimensional model according to the clock direction angle and the resection depth through a scalpel model, determining an updated three-dimensional model after myocardial resection, and marking a resection area on the updated three-dimensional model. According to the scheme, accurate preoperative planning guidance and comprehensive heart structure information presentation are provided for the myocardial rotary cutting operation, and therefore the accuracy and safety of the myocardial rotary cutting operation are improved.

The invention discloses a laser measurement system suitable for a large complex curved surface and a measurement method thereof.The measurement system comprises a control module, a positioning module, a pre-scanning module and a high-precision scanning module, and the control module comprises a high-precision scanning planning unit and a scanning analysis unit; the high-precision scanning module comprises a three-dimensional scanning platform, a universal mechanical arm and a distance measuring laser; the measurement method comprises the steps of pre-scanning to generate an initial model, scanning path planning and scanning attitude control, high-precision scanning to collect point cloud data, preprocessing and marking of the high-precision scanning point cloud data, local optimization and supplementary scanning strategy generation and high-precision scanning model generation. According to the measurement system and method, the dynamic scanning path of the high-precision scanning module is intelligently planned based on the pre-scanning result, the pre-scanning and high-precision scanning strategies are combined, high-density coverage and uniform sampling of the large complex curved surface are achieved, measurement is accurate, the anti-interference capacity is high, and the model scanning precision is high; a dynamic supplementary scanning mechanism is automatically marked and triggered in the scanning process, and the quality of scanned data is ensured.

The invention discloses a multi-view three-dimensional reconstruction method for a slender object based on curve guidance. According to the method, geometric prior and a deep learning segmentation network are fused, two-stage foreground extraction, a curve-guided motion recovery structure (SfM) and a surface optimization module capable of differential Poisson reconstruction are provided, and a set of complete processing flow is established for the problem that a slender and weak-texture object is difficult to accurately reconstruct. The method comprises the following steps: acquiring a multi-view image through a consumer-level terminal and performing frame extraction to obtain an input data set; generating a high-quality foreground mask by utilizing depth estimation and semantic segmentation; performing curve-guided SfM initialization based on the foreground skeleton curve to realize joint estimation of the camera pose and the sparse three-dimensional curve; carrying out surface geometric optimization through a micro Poisson equation and differentiable grid rendering; and designing a multi-branch loss function fusing luminosity, mask, regularization and Gaussian rendering supervision to train the model. Finally, the method can output a three-dimensional reconstruction result of the slender structure with high geometric accuracy, strong structural integrity and smooth surface, and has good robustness and application value.

The application provides a three-dimensional reconstruction method based on binocular stereo matching, and relates to the technical field of binocular stereo vision. The method combines binocular stereo matching algorithm, triangulation algorithm and surface texture mapping, and can realize surface reconstruction of scene objects in different environments. After stereo calibration is performed on a binocular camera, the internal and external parameters of the binocular camera are obtained, stereo correction is performed, then an image disparity map of the corresponding scene object is generated according to an optimized semi-global stereo matching algorithm, after the disparity information is obtained, the point cloud information converted by the object disparity is triangulated on the surface according to the triangulation algorithm, and finally three-dimensional surface reconstruction of the object is realized through the texture mapping technology. The application can reduce the influence of noise in the disparity map generation process, i.e. the semi-global stereo matching algorithm, improve the accuracy of the disparity information, and complete the three-dimensional surface reconstruction task of the object with little sacrifice of time performance.

The invention belongs to the technical field of surveying and mapping remote sensing, and discloses an automatic surveying and mapping system and method for urban and rural planning complex terrains based on image analysis. The method comprises the following steps of: acquiring multi-view heterogeneous image data of a target surveying and mapping area, performing layered decoupling processing and multi-view registration on an image, identifying spectral texture features and geometric three-dimensional features of a terrain, and performing complexity function partitioning on the surveying and mapping area according to the spectral texture features and the geometric three-dimensional features; the regional surveying and mapping demand degree is calculated by analyzing a terrain local complexity index and spatial distribution dispersion characteristics, so that a personalized sampling parameter range is screened out; constructing an optimized three-dimensional terrain model by fusing a terrain feature map, a point cloud reconstruction result and a geometric constraint attribute, predicting surveying and mapping precision, and performing comparative analysis on the predicted surveying and mapping precision and a back projection result to obtain a confidence evaluation result; the method is applied to a personalized urban and rural planning surveying and mapping scheme; according to the invention, accurate identification and personalized automatic surveying and mapping of the complex terrain are realized.

Systems and techniques for inspecting vehicles, such as cars or trucks. Some embodiments provide a method for inspecting a vehicle using a vehicle inspection system. The vehicle inspection system may include a vehicle exterior inspection system having a plurality of exterior sensor arrays, the plurality of exterior sensor arrays including respective sets of sensors oriented in different directions. The vehicle inspection system may include a vehicle undercarriage inspection system having an undercarriage sensor array including cameras configured to capture images at different angles, and a sensor configured to output vehicle position / motion data. In some embodiments, the method includes: moving the vehicle and / or vehicle inspection system; triggering capture, by the exterior sensor arrays and undercarriage sensor array, of sensor data about the vehicle; capturing the sensor data; stopping capture of the sensor data; stopping movement of the vehicle and / or vehicle inspection system; and generating a vehicle condition report.

The invention relates to a large-span box-type arch bridge damage three-dimensional visual identification method based on unmanned aerial vehicle monitoring, and the method comprises the following steps: S1, carrying out the full-section scanning of a box-type arch bridge through an unmanned aerial vehicle according to a preset flight path, and collecting a high-resolution image and three-dimensional point cloud data; s2, in combination with the collected high-resolution image and three-dimensional point cloud data, obtaining two-dimensional damage information and three-dimensional damage information by using an improved U-Net network and a PointNet + + model; s3, associating the two-dimensional damage information with the three-dimensional damage information, automatically classifying damage types, and calculating damage parameters to obtain damage detection data; s4, mapping damage detection data to the bridge BIM model, dynamically displaying damage space distribution, locally amplifying high-risk parts of a suspender anchoring area and an arch foot stress concentration area, and superposing mechanical simulation data to obtain a stress cloud picture; and S5, positioning the damage position of the bridge according to the stress nephogram and the thermodynamic diagram, then performing highlight labeling, and generating a three-dimensional visual report.

The invention provides a multi-view panoramic image indoor layout prediction method fusing semantics and planarity. The method comprises the following steps: step 1, carrying out three-dimensional semantic segmentation combined with three-dimensional point cloud reconstruction and a multi-view geometry and two-dimensional segmentation large model; step 2, obtaining a pure structured point cloud; step 3, multi-plane detection and classification; 4, extracting an initial contour; 5, generating a wall line based on plane fitting; step 6, performing multi-label graph segmentation guided by the two-dimensional contour and the wall surface; step 7, extracting a non-Manhattan ceiling line; and step 8, stretching the two-dimensional wall line along the vertical direction on the basis of the contour of the dimensional plane graph in combination with the vertical height of the room, generating an intersection line of the wall surface and the ceiling and an intersection line of the wall surface and the ground, and finally outputting a complete indoor three-dimensional vectorization layout model. According to the method, the layout integrity of the complex scene is improved, the boundary recovery precision of the occlusion area is improved compared with that of a traditional projection method, and the detection accuracy is improved by matching with multi-model RANSAC plane detection.

The invention provides an opencast coal mine data fusion method and device, electronic equipment and a storage medium, relates to the technical field of data processing, and realizes automatic access and integration of geological information, measurement data, mining parameters and safety monitoring data by constructing a unified data interaction platform and configuring a standardized interface protocol. Meanwhile, a three-dimensional real-scene model is generated based on automatic processing of field data collected by an unmanned aerial vehicle and a laser radar, loading and rendering are carried out in the platform, incremental updating is carried out on the three-dimensional real-scene model of a mining area by adopting a dynamic updating mechanism in combination with mine production real-time data, and volume change calculation is carried out based on the updated model; in addition, a cross-professional data association model is established by using an intelligent collaborative decision engine, and linkage analysis is performed according to the three-dimensional live-action model updated in real time and the safety monitoring data to generate risk behavior early warning information and output a mining plan dynamic optimization scheme, so that the problem of low data circulation efficiency in the prior art can be solved.

Disclosed in the present application are a method, apparatus and system for three-dimensional reconstruction. The method includes: acquiring a first image and a second image, wherein the first image and the second image are respectively obtained by means of different image collection devices collecting a predetermined code element image projected onto the surface of an object to be measured; acquiring a first target pixel point of each code element in the first image, and determining, in the second image, a matched pixel point for the first target pixel point of each code element in the first image; and determining the three-dimensional coordinates of the first target pixel point at least according to the predetermined pixel point coordinates of the first target pixel point in the first image and the predetermined pixel point coordinates of the matched pixel point in the second image, so as to complete three-dimensional reconstruction.

The invention discloses a light field reconstruction brain image brain region boundary extraction method based on convex hull fitting. The method comprises the steps that the p percentile of a brain image is calculated, and image cutting, contrast stretching and Gaussian filtering smooth denoising are carried out; carrying out Canny edge detection on the filtered image to obtain a binary edge image; boundary enhancement and connection are carried out through expansion and closed operation, all connected regions are extracted, and the contour with the maximum area is selected as the contour of the brain region boundary; a datum point is determined based on the contour of the brain region boundary, the polar angle and distance of each point except the datum point are calculated and sorted, and convex hull fitting is carried out through Graham Scan scanning based on the datum point and the sorted point set; converting the convex hull into a closed boundary region, and obtaining a mask image of the brain image after light field reconstruction through a mask function; according to the method, the special boundary form of the light field reconstruction image can be accurately adapted, deep learning and data training are not needed, and the generated mask has high geometric consistency.

A mobile device is capable of being carried by a user and directed at a target object. The mobile device may implement a system to dimension the target object. The system, by way of the mobile device, may image the target object to and receive a 3D image stream, including one or more frames. Each frame may include a plurality of points, where each point has an associated depth value. Based on the depth value of the plurality of points, the system, by way of the mobile device, may determine one or more dimensions of the target object.

The invention discloses an enhanced 3D (three-dimensional) surface reconstruction method based on 3D (three-dimensional) Gaussian Splitting. Global consistent depth priori is obtained through virtual stereo pair rendering; constructing a factor graph and introducing a cross-view geometry / luminosity consistency constraint to form local beam adjustment loss; the prior is used as a learnable parameter to be combined with 3DGS to be optimized, and meanwhile, the Pull loss is assisted to pull low-credibility pixels; and finally, multi-loss function end-to-end training is adopted. The method comprises the following steps of: in Tanksamp; the method has the advantages that F1 is equal to 0.58 and DTU Chamfer is equal to 0.48 mm on a Temples data set, the training time is only 20 min, compared with the prior art, geometric accuracy SOTA and speed magnitude improvement are achieved at the same time, and the method is suitable for VR / AR, robot and industrial measurement scenes.

A mobile device is capable of being carried by a user and directed at a target object. The mobile device may implement a system to dimension the target object. The system, by way of the mobile device, may image the target object to and receive a 3D image stream, including one or more frames. Each frame may include a plurality of points, where each point has an associated depth value. Based on the depth value of the plurality of points, the system, by way of the mobile device, may determine one or more dimensions of the target object.

Methods for product recommendation are provided. This includes collecting personal user data including profession, gender, size, preferences, user's apparel size, user's address, who the user's friend are, user's friends' information, users' medical records including eyeglass and contact lens prescription information; collecting vendor data including product size, product description, product location, price; and recommending vendor products that best match the users' personal data.

Disclosed herein are methods and systems for building and updating a virtual store based on a physical store. A computing system is structured to receive captured images of a physical space having a set of products. The physical space is simulated by a three-dimensional (3D) model of the physical space. A corresponding representation of the set of products is positioned relative to the 3D model. The computing system is structured to detect, based on the captured images, product information for products in the physical space and generate, based on the product information, an updated representation of the set of products based on the product information. The updated representation of the set of products corresponds to the products in the physical space as shown in the captured images.

The invention provides a three-dimensional modeling method and system based on multi-source heterogeneous data, and relates to the technical field of three-dimensional reconstruction. Multi-view point clouds, images and environment data in the same time state are obtained, and point cloud alignment precision is enhanced through feature detection and descriptor matching; an iterative nearest point algorithm is utilized to complete initial point cloud fusion to generate a three-dimensional model, position calibration is realized in combination with geometric features of environmental data, triangular meshes are divided by adopting a triangulation algorithm, surface defects of the model are repaired by adopting a hole filling algorithm, camera calibration is performed on image data, the image data are projected to the corresponding triangular meshes, and the three-dimensional model is obtained. And setting a pixel boundary threshold value to control a projection position, realizing accurate texture mapping, and defining a degradation coefficient to evaluate a local deformation condition of the three-dimensional model. According to the method, by fusing multi-source heterogeneous data and combining feature matching, position calibration and degradation coefficient evaluation, high-precision and complete three-dimensional model construction is achieved.

A method for inspecting an object includes determining a first inspection package that includes a first inspection image of the object and a first designation. The method includes determining data indicative of a second inspection package that includes a second inspection image of the object and a second designation. The method includes determining a first property of the object based on the first inspection image of the object, one or more properties maps of the object, and the first designation. The method includes determining a second property of the object based on the second inspection image of the object, the one or more properties maps of the object, and the second designation. The method includes displaying the first property and the second property or displaying data indicative of a comparison of the first property with the second property.

Systems and techniques for inspecting undercarriages of vehicles. Some embodiments provide a vehicle undercarriage inspection system. The vehicle undercarriage inspection system may include a base having a width and a depth respectively extending along first and second orthogonal directions, the width extending from a first end to a second end of the base. The vehicle undercarriage inspection system may include a sensor array coupled to the base, the sensor array including a set of cameras oriented in multiple directions. The set of cameras may include: a first camera oriented in part towards the first end of the base, a second camera oriented in part towards the second end of the base, and a third camera between the first and second. The vehicle undercarriage inspection system may include a processor configured to control the sensor array to capture images of an undercarriage of a vehicle with the set of cameras.

This invention focuses on the 3D reconstruction of dynamic crowds in large-scene videos and introduces the DyCrowd framework, which reconstructs 3D position, pose, and shape of hundreds of people from a large-scene video. Our approach addresses frequent occlusions and modeling challenges in high-density crowds through a top-down strategy. This includes pre-reconstruction, matching individual movement sequences, and multi-stage iterative optimization. During the optimization process, we introduce a group optimization method with an asynchronous motion consistency loss. This method clusters individuals with similar trajectories, using high-quality and unoccluded movements within the group to guide the recovery of occluded individuals, thereby mitigating long-term occlusion issues. Furthermore, to address the lack of ground-truth human reconstruction labels in current large-scene datasets, we introduce a virtual benchmark dataset called VirtualCrowd for dynamic crowd reconstruction in large-scene videos.

The invention provides a modeling method, related electronic equipment and a storage medium, and relates to the field of three-dimensional reconstruction. In the method, the terminal equipment can display a first interface comprising a shot picture; in response to the acquisition operation, acquiring multiple frames of images corresponding to the to-be-modeled target object, and acquiring an association relationship among the multiple frames of images; and obtaining and displaying a three-dimensional model corresponding to the target object according to the multiple frames of images and the association relationship among the multiple frames of images. Displaying a first virtual surrounding body comprising a plurality of patches when collecting multiple frames of images, and collecting a first image and changing the display effect of the patch corresponding to the first image when the terminal equipment is in a first pose; when the terminal equipment is in a second pose, a second image is collected, and the display effect of a surface patch corresponding to the second image is changed; and when the display effect of the plurality of patches is changed, obtaining the association relationship among the plurality of frames of images according to the plurality of patches. According to the method, the 3D modeling process is simplified, and the requirement for equipment hardware is low.

A method for determining a plurality of final optical parameters of a dental object in an intraoral cavity includes receiving a plurality of two-dimensional images of the dental object, reconstructing a three-dimensional model of the dental object based on the plurality of two-dimensional images, determining camera positions of the plurality of two-dimensional images relative to the three-dimensional model, and receiving a plurality of optical parameters. The method may further include determining in an iterative manner: a plurality of simulated two-dimensional images of the dental object by inputting the plurality of optical parameters and the camera positions to a differentiable renderer, and multiple loss-values based on a loss function between the plurality of simulated two-dimensional images and the received plurality of two-dimensional images by adjusting the plurality of optical parameters.

The invention belongs to the technical field of three-dimensional modeling, and particularly relates to a three-dimensional visual detection method for motorcycle tires. According to the method, a tire is fixed, multi-view point cloud scanning is carried out, and an improved SAC-IA algorithm is adopted for registration and fusion to construct a three-dimensional point cloud model; the method comprises the following steps: designing adaptive voxel grid filtering and denoising based on tire geometric prior, and after improved bilateral filtering smoothing, fitting and unifying a coordinate system through principal component analysis and least square to obtain a standardized point cloud model; and extracting four core parameters of thickness uniformity and sidewall contour curvature, inputting the four core parameters into the hybrid dynamic weight scoring network to output quality scores, and meeting the requirements of high-precision and high-efficiency quality inspection of motorcycle tires in an industrial environment.

The invention provides a method and device for measuring the thickness of a large-area coating of a curved-surface workpiece, and relates to the technical field of coating thickness measurement, and the method comprises the steps: obtaining the point cloud data of a target coating region of the curved-surface workpiece, and calculating the Gaussian curvature of each point; dividing a curved surface corresponding to the target coating area into a plurality of target curved surface patches with different curvature levels according to Gaussian curvature, and further generating a curved surface model; dividing the initial grid based on the curved surface model to obtain a plurality of initial measurement points; controlling the mechanical arm to measure the thickness of the target coating area according to each preliminary measurement point to obtain a preliminary thickness data set; adjusting grid nodes of the initial grid according to the initial thickness data set to obtain a plurality of target measurement points; controlling the mechanical arm to measure the thickness of the target coating area according to each target measuring point to obtain a target thickness data set; and generating a thickness distribution cloud picture according to the target thickness data set. According to the invention, high-precision measurement of the coating thickness of the large-area curved-surface workpiece is realized.

The present disclosure discloses a Method for Processing Three-dimensional Scanning, a three-dimensional scanning device, and a computer-readable storage medium. The method includes: projecting, by a pattern projector, a plurality of lines to a surface of an object-to-be-measured; collecting, by three cameras, two-dimensional images on the surface of the object-to-be-measured to correspondingly obtain three frames of two-dimensional images; determining matching point pairs between every two of the three frames of two-dimensional images to correspondingly obtain three sets of matching point pairs; verifying matching consistency between the matching point pairs; and performing three-dimensional reconstruction on the matching point pairs with the matching consistency to obtain three-dimensional points on the surface of the object-to-be-measured.

One variation of a method for modeling a human body includes: driving a sensor block along a path above a platform occupied by a user and recording a sequence of depth images and color images, via the sensor block, at each capture position in a sequence of capture positions along the path; compiling the set of depth images into a low-density 3D model of the user and a high-density 3D model of the user; extracting a set of color patches, from the sequence of color images, corresponding to discrete regions on a surface of the low-density 3D model of the user; projecting the set of color patches onto the surface of the low-density 3D model to generate a 3D color model of the user; and extracting a value of a dimension, projected from the low-density 3D color model onto the high-density 3D model, from the high-density 3D model.

Embodiments of the present disclosure relate to a method and system for processing intraoral scanning, an electronic device, and a medium. The method includes: obtaining a positioning mode of a scan post in response to a request for scanning an oral cavity including scan posts, scanning the oral cavity to obtain target scanning data, and performing processing based on the positioning mode of the scan post and the target scanning data to determine positioning information of the scan posts. By adopting the above technical solution, the positioning information of the scan posts can be acquired based on a scanning result, and the requirement for positioning the scan posts in an intraoral scanning scenario is further satisfied.

The utility model provides a kind of analytical modeling method of model measuring point three-dimensional coordinates and Euler angle, the utility model is suspended under crossbeam by crane by scanning workpiece, solve the deformation problem of large thin-walled part different from its working posture due to the workpiece to be scanned by gravity, scanner is installed at the end of manipulator, realize all-around scanning to workpiece, according to the distance signal of scanner to control the scanning distance between scanner and workpiece to be scanned, solve the fatigue of operator long time handheld scanner, while can stably keep scanner and workpiece to be scanned directly optimal scanning distance, the utility model can satisfy the automatic scanning of different size sheet metal parts in whole vehicle reverse engineering, solve the deformation problem of large sheet metal part different from its working posture due to gravity when being placed on traditional scanning platform, while three-dimensional coordinates and Euler angle of measuring point in whole vehicle crash test and body-in-white modal test can be obtained quickly, improve scanning efficiency, shorten reverse engineering cycle etc.