33 results about "Surfel" patented technology

A facet-based complex scene three-dimensional reconstruction method is characterized by including: firstly, obtaining intrinsic and extrinsic parameters and distortion coefficients in a depth camera and a color camera by three-dimensional calibration; secondly, calculating three-dimensional information of a complex scene by the depth camera according to the principle of triangulation; thirdly, defining a point model as a facet, with each value as an attribute of the point model including the attributes such as normal vector and radius; fourthly, subjecting each facet to surface 'internal, external and intersection' relation test with another facet; fifthly, obtaining confidence neighborhood of the facets which are judged intersected, calculating boundaries of the most adjacent facets on the point model, and resampling the facets in the confidence neighborhood; sixthly, three-dimensionally drawing the complex scene according to results of Boolean operation in the step 5. The facet-based complex scene three-dimensional reconstruction method is highly timely and available for effectively extracting three-dimensional point cloud data of the complex scene.

The invention belongs to the field of image processing, and provides a semantic segmentation method and system for removing dynamic objects. The semantic segmentation method comprises the following steps: acquiring a scene 3D point cloud, projecting the scene 3D point cloud into a 2D depth map, calculating normal vector information of the point cloud, and constructing a surface element map with a loop; carrying out point cloud semantic segmentation on the surface element map with the loop, constructing a semantic point cloud map, mapping the semantic point cloud map to a 3D point cloud map, and removing edge shadows and point cloud discretization phenomena; and removing the dynamic object by utilizing semantic point cloud information in the semantic point cloud map, and adding semantic iteration nearest point constraint to obtain an optimized semantic point cloud map.

The invention discloses a remote sensing inversion method for surface high-resolution spectral information based on irradiance, comprising the following steps: S1, three-dimensional scene reconstruction; S2, angle normalization; S3, setting sensors and light sources; S4, calculating shape Factor; S5, set temperature and initial value of reflectivity; S6, output simulated image through GPU fast radiosity method; S7, threshold judgment; S8, optimize temperature and reflectivity of surface element; S9, end process and output result. The present invention fully considers the influence of multiple scattering on the ground surface, so that the temperature and reflectance results of the inversion are both stable and accurate; the strategy based on GPU instead of CPU makes the present invention more practical for inversion, and this method is expected to become Effective tool for high-resolution surface temperature and albedo production from drones or manned aircraft.

The invention discloses a dynamic three-dimensional display method of electromagnetic waves suitable for complex curved surfaces, comprising the following steps: S1. Reading the point list of the complex curved surface model from the nastran grid file [P] N Join list with points to form list of triangle surfels [Tri] M ; Among them, N is the number of points, M is the number of triangles, initialize a data table [Tbl] whose dimension is N×8 N,8 ; S2. Calculate point coordinates and unit normal vectors for each surface element in the surface element list, and accordingly to the data table [Tbl] N,8 Filling; S3. To the filled data table [Tbl] N,8 Process each line of data in and complete the arrangement of visualized data; S4. Perform dynamic visualization processing on the electromagnetic waves on the complex curved surface to realize the dynamic three-dimensional display of electromagnetic waves. The invention can realize the fast and slow effects of electromagnetic fluctuations, the fluctuation amplitude, and the adjustment of electromagnetic fluctuations forward and backward with time, and avoid the problem that the peak or trough reverses due to the connection sequence of points in different grid models.

The invention discloses a space target imaging simulation surface element blanking method, relates to a space target imaging simulation surface element blanking method, and belongs to the technical field of target imaging simulation. In order to solve the problem of long time and low efficiency in the method of traversing all the facets to judge the occlusion of the facet under the condition of a large number of micro-facets. A method for blanking surface elements of space object imaging simulation according to the present invention includes a surface element blanking part to the sun light source and a camera blanking part, and the surface element blanking part to the sun light source refers to judging each micro-facet on the surface of the space object Whether the element can receive sunlight, the part of the face element to the camera blanking refers to judging the visibility of each micro-facet element of the space object to the camera. Intersect the surface elements visible to the sun light source and the surface elements visible to the camera to get all the surface elements visible to both the sun light source and the camera. The present invention is suitable for surface element blanking in space object imaging simulation.

The invention relates to a dynamic simulation and velocity estimation method of a complex number domain structured SAR ship target, which belongs to the field of SAR image processing. The invention aims to solve the problem that there is a lack of a large number of simulation samples when doing SAR ship target simulation, and the existing SAR ship target simulation method cannot obtain accurate ship target SAR images, resulting in the inability to obtain accurate ship target problems. The method of the present invention comprises: obtaining a 3D model of a ship, preprocessing the 3D model of a ship, dividing the preprocessed 3D model of a ship into a plurality of triangular surface elements; performing ray tracing according to set radar parameters, and obtaining a ship Spatial coordinates of target scattering points; ship target imaging; input training samples into AlexNet network to obtain a trained AlexNet pre-training model, input the samples to be tested into the trained AlexNet pre-training model, and calculate the speed estimation result of AlexNet network complex number domain. The invention is used for dynamic simulation and speed estimation of SAR ship target.

The invention discloses a pre-evaluation system for accelerating computation of target radar scattering cross section by graphical electromagnetic computation through index information. The system comprises a surface element modeling module (1), an index information compilation module (2), an entity display module (3), an index information unscrambling module (4), a geometric information analysismodule (5) and a target radar scattering cross section acquisition module (6). The surface elements of a target model are numbered, an index list of the surface elements is established, the surface element index and the surface element color are subject to one to one correspondence, the background color is set as black, components of three primary colors namely red, green and blue are zero, the surface element blanking is finished by a GPU, the corresponding surface element index is obtained by the color of pixels so as to obtain normal vectors and depth values of the pixels. The whole process needs no illumination, the surface normal vector is pre-obtained in a drawing process of the surface elements, a color value is read only for once, and twice memory exchange is executed, which is one computation less than the normal graphical electromagnetic computation, thus shortening the computation time.

The invention discloses a microwave correlated imaging simulation method based on an analysis surface element. The method comprises the steps: firstly constructing a target 3D model scene, obtaining the space information of a corresponding target scene according to observation parameters, calculating a single-scattering coefficient and a multi-scattering coefficient of the target scene, and obtaining the scattering coefficient distribution of the observation scene; secondly calculating an echo signal of the target scene according to a radiation field of a random antenna wave beam in an imaging plane; finally carrying out the correlating of the radiation field and the echo signal, and obtaining a simulation image result. The method achieves a quick simulation process in correlation imaging under the condition of the random radiation field, gives a solution for distributed scene correlation imaging high-efficiency simulation, can deepen the understanding of the imaging mechanism and scattering mechanism in a whole correlated imaging link, provides support for the design and index argument of the design of a correlated imaging radar system, also can be used for subsequent correlated imaging evaluation and image interpretation, and is high in practical value.

The invention discloses a method for extracting building wall and window information in a SAR image. The specific steps are: 1) using a point feature extraction method to obtain building bright spots in the SAR image; 2) using Radon transform to obtain the corresponding building bright spots in the SAR image Azimuth spacing; 3) Use the maximum texture correlation method to obtain the distance spacing corresponding to the bright spots of the building in the SAR image; 4) Use the Radon transform projection variance method to obtain the azimuth angle of the building wall in the SAR image; 5) Based on the SAR image building The Delaunay triangulation network of the highlight uses the triangular surface element matching method to extract the exact number and spatial distribution of windows on the building wall. The invention performs a series of image processing processes on the SAR image, extracts the number of windows on the wall of the building and their distribution, and realizes the extraction of the fine structure of the target wall of the building, which can satisfy the fine recognition and interpretation of the artificial target in the SAR image. It has great application prospects in areas such as city rules and land monitoring.

The invention discloses a kind of point fast matching method applicable to the method of moments, comprising the following steps: S1. read the point list of the model from the nastran grid file [P n ] Nnode Concatenate list with points to form list of surfels[Elem] Nelem , and generate a model view; S2. The user operates the model view to perform rotation, translation, zoom in / out, and restore operations, modify the coordinate axis, coordinate range, and upper left corner coordinates of the current screen display area, and trigger the initialization process before point matching; S3 .The user turns on the point matching mode of the mouse on the model; S4. The user circulates and matches N points in the model to form a list of matching points [P track ] N ; S5. The user closes the point matching mode; S6. Repeat steps S2-S5 until the quick matching of all points in the model is completed, and all matching points are collected to form [P track ] N′ . The invention effectively improves the speed and accuracy of point matching in surface element grids.

The present invention relates to a pixel subdivision load transfer method and system. The pixel subdivision load transfer method comprises: obtaining surface pressure distribution data, the data being in the form of surface elements and corresponding pressure; subdividing surface element pixels into subdivision surface elements which are small enough; allocating the subdivision surface elements to corresponding structure bearing points; integrating the subdivision surface elements allocated to the same structure bearing point into concentrated force and moment, and transferring the concentrated force and moment to the structure bearing point; comparing load distribution before the transfer with load distribution after the transfer to obtain a difference, and determining whether the difference is less than given tolerance; if the difference is less than the given tolerance, obtaining a final load; and if the difference is greater than or equal to the given tolerance, continuing performing pixel subdivision on the surface elements, and repeating the subsequent steps until the final load is obtained. The pixel subdivision load transfer method and system of the present invention can achieve the following beneficial technical effects: transferring a pneumatic surface load onto a structure node, and achieving conservation before and after the transfer, and a precise match of load distribution.

The invention discloses a near-field electromagnetic scattering simulation method for ultra-electric large-scale targets: import a model file in STL format, read the relevant information of each triangular surface element that constitutes a radar target; input parameters that need to be calculated; judge the surface Whether the element is illuminated by the incident wave, and mark the illuminated surface element; for each triangular surface element marked as illuminated, calculate its surface current and magnetic current: solve each triangular surface element marked as illuminated The resulting scattered field will be the scattered field E of all triangular surfels marked as illuminated sn After all are solved, according to the principle of vector superposition, add them all up to get the total scattering field; get the RCS value σ of the radar target in the near field according to the following formula 0 , and output the result: the present invention fills the gap in the field of target near-field RCS simulation algorithm, especially the electromagnetic simulation processing method of different surface elements under near-field scattering conditions, which is more suitable for actual engineering scenarios.

The invention discloses an automatic separation method for a ground model and a non-ground model for a three-dimensional model of a real scene. The present invention simulates the free fall of the cloth on the inverted three-dimensional model of the real scene by using the particle physics model with the distance constraint for the surface element, and determines the final form of the cloth through the collision detection between the particle and the three-dimensional model of the real scene, which is used as an approximation of the ground model, The ground model and the non-ground model are locally distinguished by comparing the spatial relationship between the particle and the surfel of the 3D model of the real scene. The invention has the advantages of high efficiency, high robustness and high precision, and can be used for the separation of the ground model and the non-ground model of the real scene three-dimensional model of urban, woodland, rural and other scenes and any surface element density.

A method and system for evaluating bin uniformity of a 3D seismic observation system. The method includes: 1) inputting the azimuth-offset data pair of each bin in the 3D seismic observation system; 2) mapping the azimuth-offset data pair of each bin on the respective bin 3) Set a standard bin two-dimensional space with uniform azimuth and offset, and map its azimuth-offset data pair into the standard bin two-dimensional space; 4) For one of the bins, Utilize the standard bin two-dimensional space in step 3) and the data calculation correlation of the bin two-dimensional space of the bin; 5) traverse all bins of the three-dimensional seismic observation system, and calculate the correlation of each bin . The present invention quantitatively calculates the correlation degree of each bin azimuth and offset two-dimensional space with the standard bin azimuth and offset two-dimensional space distribution, and quantitatively evaluates the bin azimuth and offset of the observation system Distribution uniformity, and can give evaluation results through intuitive data graphics.