36 results about "Surfel" patented technology

The invention relates to the field of three-dimensional reconstruction, and discloses a three-dimensional scene online reconstruction method, device and equipment and a storage medium, and the method is used for realizing the online reconstruction of a three-dimensional scene. The method comprises the following steps: acquiring and processing multiple frames of ordered scene RGB images to obtain sparse point clouds of each frame of scene RGB image under a camera coordinate system, and converting the sparse point clouds into sparse point clouds under a world coordinate system; initializing a sparse point cloud under a world coordinate system to obtain a local Gaussian surface element of each frame of scene RGB image; based on a fusion mechanism of a gating loop unit, reading global features of sparse point clouds of each frame of scene RGB image in a world coordinate system frame by frame from the first frame, and fusing the global features of the scene RGB image of the previous frame with the local Gaussian surface element of the scene RGB image of the current frame to generate a global Gaussian surface element of the scene RGB image of the current frame; and rendering the global Gaussian surface element of each frame of scene RGB image from a plurality of visual angles, and generating a scene reconstruction surface by using a Poisson reconstruction algorithm.

The invention discloses a single-view human body three-dimensional reconstruction method based on Gaussian surface elements, and the method comprises the steps: collecting image data from a plurality of human body data sets, carrying out the preprocessing through employing a targeted preprocessing strategy according to the data characteristics corresponding to each image data, and obtaining a processed human body data set; constructing a human body 3D Gaussian attribute prediction model, wherein the human body 3D Gaussian attribute prediction model comprises a hierarchical feature extraction module, a feature fusion network and a 3D Gaussian function decoder; training the human body 3D Gaussian attribute prediction model based on the processed human body data set to obtain a trained human body 3D Gaussian attribute prediction model; and performing three-dimensional reconstruction on an input human body picture based on the trained human body 3D Gaussian attribute prediction model, and outputting a reconstructed human body three-dimensional image. According to the method, the manufacturing cost of the virtual digital human is greatly reduced, and the efficiency and accuracy of constructing the dynamic human body model are improved.

The invention provides an underwater multi-mode intelligent sensing method for a marine ranch, and the method comprises the following steps: introducing a sonar sensing mode based on a YOLO underwater multi-mode detection module, and constructing a dynamic self-adaptive multi-mode sensing system; semantic SLAM and dense coarse mapping are carried out; the semantic SLAM and dense coarse mapping process comprises the steps of preprocessing, initialization and joint optimization of visual-inertial data, dense mapping based on surface elements and semantic fusion mapping; physically-driven 3D GS fine reconstruction is carried out, wherein the 3D GS fine reconstruction comprises initial 3D GS modeling and underwater medium estimation and compensation; and outputting the multi-scale semantic map. According to the invention, the physical imaging model is embedded into the 3DGS rendering process, and the attenuation coefficient and the scattering parameter are estimated through iteration, so that the color fidelity of the reconstructed scene is improved. And compared with the traditional method, the reconstruction loss of the compensated image and the original image is obviously reduced.

An example method includes obtaining a LIDAR scan including a plurality of LIDAR points. The LIDAR scan being descriptive of an environment of an autonomous vehicle. The method includes obtaining a surfel map descriptive of the environment. The surfel map includes a plurality of surfels. The method includes classifying each of the plurality of LIDAR points as an explained LIDAR point or an unexplained LIDAR point based on an alignment of the LIDAR scan to the surfel map, wherein an explained LIDAR point is a LIDAR point that matches a surfel in the surfel map. The method includes shifting the LIDAR scan along a longitudinal direction for a plurality of iterations. The method includes determining a longitudinal bias for each of the plurality of iterations. The method includes performing an operation for the autonomous vehicle based on the longitudinal bias.

The invention discloses a text-to-three-dimensional surface generation method and system based on a two-dimensional Gaussian surface element, and belongs to the technical field of three-dimensional reconstruction, and the method comprises the steps: representing a three-dimensional object surface as a Gaussian surface element on a local tangent plane; receiving a text description, generating normal diagrams and texture diagrams of a plurality of visual angles based on a pre-training text-to-image diffusion model, and generating a preliminary background mask at the same time; determining the center of sphere and the maximum radius according to the extrinsic parameters of the camera, and randomly generating Gaussian surface elements in the spherical range; removing the surface elements falling in the preliminary background mask area; on the basis of a two-dimensional Gaussian dot drawing renderer, the normal graph and the texture graph are rendered from multiple perspectives, and curvature consistency regular loss and surface convergence constraint loss are calculated; geometric and texture parameters of Gaussian surface elements are optimized through back propagation, iteration is carried out until convergence, and a final three-dimensional model is output. According to the method, the high-fidelity three-dimensional surface can be generated from the text without SfM (Structural Recovery Motion) initialization, and the method has relatively high geometric consistency and generation speed.

The present invention discloses a method for recording and displaying photopolymers based on a computer-generated holographic bin algorithm. S1: After a beam of light emitted by a femtosecond pulse laser is expanded and collimated, the pulse laser is focused on a photopolymer material through an objective lens, causing the refractive index of the photopolymer to change. The position of the photopolymer is controlled by controlling a translation stage to achieve three-dimensional recording of two pieces of material. S2: The complex amplitude distribution of the object is calculated through the computer-generated holographic bin algorithm and the two pieces of material are encoded. The exposure time and other parameters are controlled by a computer, and the femtosecond laser is used to write the calculated and encoded phase information diagram on the photopolymer. S3: After recording, the second beam splitter and the objective lens in front of the recording position are removed, and a plane wave is used to simultaneously irradiate the two pieces of engraved photopolymer and combine the beams to obtain a reproduced image of the three-dimensional object. The recording and display of the three-dimensional object can be achieved.

The present invention relates to the field of three-dimensional reconstruction, and discloses an online three-dimensional scene reconstruction method, apparatus, device and storage medium. The method is used to realize the online reconstruction of a three-dimensional scene. The method includes: acquiring and processing multiple frames of ordered scene RGB images to obtain sparse point clouds of each frame of scene RGB image in the camera coordinate system, and converting them into sparse point clouds in the world coordinate system; initializing the sparse point clouds in the world coordinate system to obtain local Gaussian surface elements of each frame of scene RGB image; based on the fusion mechanism of the gated recurrent unit, reading the global features of the sparse point clouds of each frame of scene RGB image in the world coordinate system frame by frame starting from the first frame, and fusing the global features of the previous frame of scene RGB image with the local Gaussian surface element of the current frame of scene RGB image to generate the global Gaussian surface element of the current frame of scene RGB image; rendering the global Gaussian surface element of each frame of scene RGB image from multiple perspectives, and using the Poisson reconstruction algorithm to generate the scene reconstruction surface.

The invention discloses a real-time three-dimensional reconstruction method based on a dynamic target in a video, and the method comprises the following steps: S1, collecting a video sequence, and extracting an image frame and a corresponding depth image; s2, executing target extraction and deep processing to generate an initial surface element set; s3, an inter-frame registration graph structure is constructed based on SurfelWarp, and non-rigid registration is executed; s4, performing surface element activation detection according to depth change and boundary features, and inserting a newly added surface element; s5, the space distance between the surface elements and the normal included angle are judged, and surface element combination is executed; s6, counting registration times and shielding states, and removing the surface elements which do not meet the continuation conditions; s7, adjusting the surface element density according to the registration error; and S8, outputting the updated surface element set and the registration graph structure, and generating a three-dimensional reconstruction model sequence. According to the invention, real-time three-dimensional modeling for a dynamic target is realized.

The data processing method can be applied to a communication perception integrated system, and the method comprises the steps that a first device sends a reconstruction instruction and a reconstruction quality requirement to at least one second device, the reconstruction instruction is used for instructing point cloud data of a perception target surface to be reconstructed into a surface element, and the reconstruction quality requirement is used for instructing a quality parameter of the reconstructed surface element; and the second device reconstructs the target point cloud data into a surface element according to the reconstruction indication and the reconstruction quality requirement, and then sends the surface element to the first device. Because the surface elements are formed by specific vertexes according to a certain topological connection relation, a grid formed by the surface elements reduces many points relative to target point cloud data, and the reconstructed surface elements are constrained by quality parameters of the reconstructed surface elements, so that actual scene requirements can be met. In this way, under the condition that the actual scene requirement is met, compared with point cloud data sending, the second device sends the surface element to the first device, the data transmission amount is reduced, and the data storage amount of the first device is also reduced.

This paper describes a method for compressing both untracked and tracked meshes using a projection-based approach, leveraging tools and syntax already developed for projection-based point cloud compression. Similar to the V‑PCC approach, the mesh is partitioned into surface surfels, with the difference that the partitioning respects the connectivity of the mesh. Each surface surfel (or 3D surfel) is then projected onto a 2D surfel, resulting in a triangular surface sampling similar to common rasterization methods used in computer graphics in the case of a mesh. For each surfel, the positions of the projected vertices are saved in a list along with the connectivity of these vertices. The sampled surface is similar to a point cloud and is encoded using the same methods used for point cloud compression. Additionally, a list of vertices and connectivity is encoded per surfel and the data is sent along with the encoded point cloud data.

The application discloses a single-view human body three-dimensional reconstruction method based on a Gaussian surface element, and comprises the following steps: collecting image data from multiple human body data sets, pre-processing the image data according to the corresponding data characteristics of the image data by adopting a targeted preprocessing strategy, and obtaining a processed human body data set; constructing a human body 3D Gaussian attribute prediction model, wherein the human body 3D Gaussian attribute prediction model comprises a hierarchical feature extraction module, a feature fusion network and a 3D Gaussian function decoder; training the human body 3D Gaussian attribute prediction model based on the processed human body data set, and obtaining a trained human body 3D Gaussian attribute prediction model; and performing three-dimensional reconstruction on an input human body picture based on the trained human body 3D Gaussian attribute prediction model, and outputting a reconstructed human body three-dimensional image. The application greatly reduces the cost of virtual digital human production, and improves the efficiency and accuracy of constructing a dynamic human body model.

The present invention discloses a method for real-time 3D reconstruction based on dynamic targets in a video, comprising the following steps: S1, capturing a video sequence and extracting image frames and corresponding depth images; S2, performing target extraction and depth processing to generate an initial facet set; S3, constructing an inter-frame registration graph structure based on SurfelWarp and performing non-rigid registration; S4, performing facet activation detection based on depth changes and boundary features and inserting new facets; S5, determining the spatial distance and normal angle between facets and performing facet merging; S6, counting the number of registrations and occlusion status, and eliminating facets that do not meet the survival condition; S7, adjusting the facet density based on the registration error; S8, outputting the updated facet set and registration graph structure to generate a 3D reconstruction model sequence. The present invention realizes real-time 3D modeling for dynamic targets.

This invention discloses a fuze target matching and adaptive detonation height control method, belonging to the field of fuze electronic countermeasures technology. The implementation method of this invention is as follows: The relative velocity is obtained based on detection information, and then the optimal activation delay τ is calculated based on the target's two-dimensional image. opt Optimal startup delay τ opt The goal is to minimize the total detonation deviation of each surface element of the target. Combining the near-field two-dimensional imaging results of the target with the fuze's target, the target centroid is estimated by weighting the grayscale of the near-field extended target's two-dimensional image, and the detonation height is designed based on the target centroid. A convolutional neural network for target matching and recognition is constructed. Based on the probability of target classification output by the convolutional network, a fuze detonation height decision risk factor is designed, and a minimum-risk detonation height decision model is established. The optimal detonation height is adaptively selected according to the minimum-risk principle, thereby achieving adaptive detonation height control for known or unknown targets in complex environments and effectively improving the fuze's accuracy in estimating the detonation height of near-field extended targets.

The invention discloses a three-dimensional object local confrontation attack method based on a maximum aggregation area sparse strategy, and in a physical 3D environment, the existing target detection-oriented confrontation attack (3D-AE) mainly faces the following challenges: in order to achieve a maximum attack effect, relatively large and dispersed confrontation patches are often selected, and the confrontation effect is poor. Therefore, the confrontation patch is too obvious, and the visual concealment is reduced. A better strategy is how to utilize as small as possible, and the attack effect of the aggregated adversarial patches is maximized. In order to maximize aggregation of adversarial patches, an aggregation degree regular term is designed to constrain a shielding aggregation matrix obtained based on a surface element adjacency relation; in order to ensure minimization of an adversarial patch area, sparse regularization is designed, so that shielding weights tend to be distributed in a U shape, and extreme values are limited. Expanded data of visual angle changes are obtained through neural rendering, and a universal important decision-making area under the target multi-angle condition is positioned through suppression model detection.

The application provides a three-dimensional mesh generation method based on patch autoregressive compression and an electronic device. The three-dimensional mesh generation method based on patch autoregressive compression comprises the following steps: converting an input three-dimensional shape representation into an ordered triangular patch sequence, wherein each triangular patch is defined by the coordinates of its three vertices; using a shape encoder to encode the input three-dimensional shape representation into a latent representation; using an autoregressive decoder to generate each triangular patch in the ordered triangular patch sequence one by one in sequence, with the latent representation as a condition, thereby generating a three-dimensional mesh. By representing a three-dimensional mesh as an ordered patch sequence and performing autoregressive generation, the modeling granularity is improved from the bottom vertex coordinates to the high-level triangular patch. The model can directly learn and generate the topological structure of the mesh, significantly improving the ability to capture the geometric structure and topological relationship of the mesh, thereby achieving the purpose of improving the generation quality and efficiency.

The invention discloses a novel representing, drawing and reconstructing method of a radiation field. According to the method, Gaussian enhancement surface elements are used as representation of a radiation field, and for a given view angle, rasterization is firstly carried out on opaque surface elements through a standard graph pipeline to generate color and depth maps; and then projecting the three-dimensional gauss to a screen space and executing a depth test, and weighting and accumulating gauss colors on each pixel of a color image rendered by the surface element according to the opacity of the gauss to obtain a final rendering result. The invention further provides an optimization process from coarse granularity to fine granularity. Compared with the prior art, the method has the advantages that high-quality new-view-angle images and videos meeting visual consistency can be synthesized at an extremely high rendering speed. In addition, the novel representation provided by the invention is combined with the existing method, so that anti-aliasing rendering can be realized, the rendering speed is further improved, the storage overhead is reduced, and more accurate scene geometry can be reconstructed by replacing three-dimensional gauss with two-dimensional gauss.

The application belongs to the technical field of image processing, and particularly relates to an MES process management and control method and system for circuit breaker production, which comprises the following steps: constructing a process feature dictionary model library, including a good product gradient dictionary and a defect gradient dictionary; receiving a verification instruction sent by a manufacturing execution system, performing hierarchical anchoring to locate and extract a real-time key feature area; extracting a surface element in the real-time key feature area, calculating good product data energy and defect data energy according to a local gradient feature vector of the surface element; adjusting spatial smoothing energy according to label consistency and local edge direction of adjacent surface elements, defining and optimizing a system total energy function containing data energy and spatial smoothing energy, and generating a globally optimal label field; generating a process defect score according to the globally optimal label field, and feeding back to the manufacturing execution system to perform process management and control closed loop. The application improves the accuracy of defect detection and realizes automatic management and control of the production process.

The invention discloses a data processing method, device and equipment and a readable storage medium, and the method comprises the steps: determining surface elements hit by M screen probes in a virtual scene space as to-be-propagated surface elements, and obtaining sampling light rays emitted by the to-be-propagated surface elements in P light ray directions in the virtual scene space; and determining illumination information stored in the surface element hit by the sampling light rays in the P light ray directions in the virtual scene space as indirect illumination information collected by the sampling light rays in the P light ray directions in the virtual scene space, thereby generating surface element mixed illumination information of the surface element to be propagated, and probe mixed illumination information of the M screen probes is generated based on the surface element mixed illumination information and the to-be-propagated surface element, and illumination rendering is performed on the screen image based on the probe mixed illumination information. By adopting the method and the device, indirect illumination information generated by multiple reflections of surface elements in a virtual scene space can be collected, and the reality sense of screen picture illumination rendering and the user experience are improved.

This invention provides a semantic mapping method and system based on visual SLAM and YOLOv5, achieving positioning optimization and globally consistent semantic mapping. The design first uses line feature optimization and weighted optimization strategies to obtain more accurate camera poses to improve mapping quality. It then integrates the YOLOv5 semantic segmentation network to obtain semantic information. A semantic map construction method combining facets and supersets is used to address semantic inconsistencies during the mapping process, ultimately constructing a globally consistent semantic map. This invention enables real-time tracking, mapping, and back-end optimization, resulting in high-precision pose estimation and globally consistent semantic map construction.

The present invention belongs to the field of electromagnetic field numerical calculation, and particularly relates to a method and device for generating FDTD conformal grids. An automatic FDTD conformal grid generation method based on 3D CAD includes the following steps: generating a 3D model based on triangular facets using CAD software, and dividing FDTD grid lines on the 3D model; performing longitudinal ray scanning and transverse plane scanning on the 3D model according to the FDTD grid lines, and calculating the geometric information and material information of each boundary cell; substituting the geometric information and material information of the boundary cells into the 3D-CP-EP algorithm formula set, and simultaneously calculating the 3D-CP-EP conformal grid parameters of the three direction components to generate an FDTD conformal grid model. The grid generation rate of the present invention is increased by a large margin, and the computational complexity of the present invention is smaller compared with traditional conformal grid generators. The grid quality also reaches the grid quality of mainstream FDTD simulation software. And with the improvement of resolution, the present invention shows better convergence.

The invention relates to the technical field of computers, in particular to a rapid simulation method and device for a three-dimensional ground temperature field. The method comprises the following steps: firstly, obtaining geometric parameters and material information of a ground scene, then determining a geometric model of the ground scene according to the geometric parameters, and endowing the geometric model with the material information, thereby obtaining a three-dimensional model of the ground scene. Then, material information, shadow information and a normal vector of each vertex are obtained, and materials, shadows and normal vectors of all pixel points in the surface element are obtained through GPU automatic interpolation; and searching in a preset temperature table according to the information so as to obtain temperature information of all vertexes and pixel points in the three-dimensional model. And endowing the temperature information of each vertex and the temperature information of all pixel points in each surface element into the three-dimensional model to obtain a three-dimensional ground temperature field. Therefore, the technical problem of how to improve the simulation speed of the three-dimensional ground temperature field can be solved.

The disclosed technology provides solutions for improving virtual object representation in a simulated (or virtual) environment and in particular, provides solutions for rendering surfel representations of real-world objects into a simulated environment using point cloud data. In some aspects, the disclosed technology includes a process for receiving a plurality of Light Detection and Ranging (LiDAR) points corresponding with a real-world object, determining, for each of the plurality of LiDAR points, an image texture based on the real-world object, and generating a surfel object representing the real-world object. In some aspects, the process further includes steps for inserting the surfel object into a simulated three-dimensional (3D) environment. Systems and machine-readable media are also provided.

A method and apparatus for performing quantum error correction using an automorphism code related to the honeycomb code. The automorphism code is based on Kramers-Wannier (KW) duality. For embodiments on a hexagonal lattice using three repeated time steps, ⅓ of the pixels are active in a given time steps. In a given time step r, a KW circuit is applied to plaquettes labeled r mod 3, transferring quantum information from the active qubits at the beginning to a new set of active qubits at the end of the time step. Each of the three plaquette types is associated with one stabilizer either given by a product of six Z operators or three X operators. The KW circuit maps the product of X operators to the product of Z operators and vice-versa. The stabilizer group of the superlattice toric code changes every round.

The application provides an underwater multi-modal intelligent sensing method for a mariculture, comprising the following steps: an underwater multi-modal detection module based on YOLO, introducing a sonar sensing mode, and constructing a dynamic self-adaptive multi-modal sensing system; performing semantic SLAM and dense coarse mapping; the process of performing semantic SLAM and dense coarse mapping comprises: preprocessing, initialization and joint optimization of visual-inertial data, dense mapping based on a surface element, and semantic fusion mapping; performing physically driven 3DGS fine reconstruction, comprising: initial 3DGS modeling and underwater medium estimation and compensation; and outputting a multi-scale semantic map. The application embeds a physical imaging model into a 3DGS rendering process, iteratively estimates attenuation coefficients and scattering parameters, and improves the color fidelity of the reconstructed scene. The reconstruction loss of the compensated image and the original image is significantly reduced compared with the traditional method.

The invention belongs to the technical field of image processing, and particularly relates to an MES process management and control method and system for circuit breaker production, and the method comprises the steps: constructing a process feature dictionary model library which comprises a good product gradient dictionary and a defect gradient dictionary; receiving a verification instruction sent by the manufacturing execution system, and executing hierarchical anchoring to position and extract a real-time key feature region; extracting surface elements in the real-time key feature region, and calculating good product data energy and defect data energy according to local gradient feature vectors of the surface elements; adjusting spatial smoothing energy according to label consistency of adjacent surface elements and a local edge direction, defining and optimizing a system total energy function containing data energy and the spatial smoothing energy, and generating a global optimal label field; and generating a process defect score according to the global optimal label field, and feeding back the score to the manufacturing execution system to execute a process control closed loop. According to the invention, the accuracy of defect detection is improved, and automatic control of the production process is realized.

The invention relates to a data processing method and device, a graphics processor and equipment, and relates to the technical field of data processing. The method comprises the following steps: acquiring first data information corresponding to a ray tracing task, wherein the first data information comprises at least one of the following data: multiple pieces of ray data with the same precision, multiple pieces of bounding box data with the same precision and multiple pieces of surface element data with the same precision; according to the precision configuration information, the precision of at least one piece of data in the first data information is adjusted to obtain second data information, and the second data information comprises at least one of the following data: multiple pieces of light data with different precision, multiple pieces of bounding box data with different precision and multiple pieces of surface element data with different precision; and performing an intersection test based on the second data information to obtain an intersection test result. In this way, the problems that in the prior art, the light tracing calculation amount is large, and power consumption is large can be solved.

The application discloses a method for calculating double-aircraft relative dynamic LRCS and a control method thereof, and combines rough surface scattering theory and computer graphics, and establishes an LRCS calculation model by using the Kirchhoff method; geometric modeling is carried out on the aircraft target, the model target surface is meshed, is decomposed into a plurality of triangular surface elements, and is coded; the triangular surface elements are subdivided, a conversion relationship between a target coordinate system and a local coordinate system is established, and coordinates are determined; a coordinate system of an observation point and a target to be observed is established, the target aircraft is blanked, effective surface element data is acquired, the value of the LRCS of each surface element is obtained, and the value of the total target LRCS at the current time is obtained through algebraic summation. According to the working principle of the laser radar, the 3D modeling and simulation are combined with the laser radar target scattering theory, and the dynamic laser scattering cross section of the other party under the mutual visual angle between two high-speed flying carriers can be researched.

The invention provides a surface element attribute evaluation method and device, and the method comprises the following steps: 1, determining n distance segments, and enabling the intervals between the distance segments to be equal; 2, aiming at each surface element of the seismic data, storing offset data in the surface element into an offset array; 3, for each surface element, counting the number of offset data included in each distance section; step 4, calculating Simpson coefficients of each surface element according to the number of the offset data; and 5, generating a Simpson coefficient distribution diagram and a histogram according to the Simpson coefficient of each surface element. According to the surface element attribute evaluation method, surface element attribute evaluation is carried out based on the Simpson coefficient, the occurrence probability of the offset in the surface element in different ranges can be quantitatively evaluated, and a more effective analysis means is provided for acquisition designers.

This invention discloses a method for displaying MR virtual images based on generative adversarial networks (GANs), comprising the following steps: constructing a display residual ledger, collecting anchor points, visibility, illumination, inter-frame residuals, and pose compensation; setting phase, frequency, visibility, frequency division, wraparound, and feedback nodes in the SIREN generator; modulating the phase and frequency terms with the anchor point phase residuals to form a phase-locked implicit state; generating an occlusion subtraction gate with the visibility residuals to correct inter-layer propagation; subtracting the illumination band residuals with frequency division to correct brightness and boundaries; wrapping the phase with the inter-frame phase residuals to form a continuous phase implicit state; decoding the surface element display record, identifying and feeding back the residuals, and stitching together to output the MR virtual image. This invention relates to the field of mixed reality display and artificial intelligence image generation technology, achieving stable display of MR virtual images and improving anchor point accuracy, occlusion fusion, and inter-frame continuity.