44 results about "Projection View" patented technology

The invention discloses a large language model robustness visual diagnosis method based on multi-dimensional features and adversarial attacks. The method aims to break through a mode that traditional evaluation only depends on a single aggregation index, and a multi-dimensional text feature exploration system covering vocabularies, syntax, semantics and a structural layer is constructed, and a large-scale antagonism disturbance mechanism and a task self-adaptive quantification strategy are combined. And generating structured feature-adversarial instruction-robustness diagnosis data comprising the cue word to be evaluated and the corpus. On the basis, an interactive visual analysis system is constructed, and through bidirectional linkage of a feature statistical view and a semantic projection view, a user is supported to realize progressive exploration from macroscopic feature screening to microscopic semantic attribution under the double view angles of cue words and corpora, so that a root cause causing the fragility of the model is deeply diagnosed. According to the method, the key feature combination influencing the stability of the model can be identified, so that a basis is provided for directional optimization of the model, and the diagnosis depth of robustness evaluation is improved.

Provided in the embodiments of the present disclosure are an object classification method and apparatus, and a computer program product, a device and a storage medium. The method comprises: acquiring a three-dimensional model of an object to be classified, wherein the three-dimensional model carries texture information; projecting the three-dimensional model according to a plurality of preset projection viewing angles, so as to obtain a plurality of frames of RGB images; selecting a target RGB image from among the plurality of frames of RGB images, wherein the information amount of valid information in the target RGB image meets a preset condition, and the valid information is information which can be used for identifying the material of said object; and performing feature extraction on the target RGB image, and determining, on the basis of an extracted feature, a material category to which said object belongs. In this way, the material category of the object to be classified can be distinguished, and when the material category is distinguished, a relatively accurate classification result can also be obtained in the case of a relatively small amount of calculation.

The invention discloses a virtual reality roaming method and system based on a multi-view panoramic image model, and belongs to the technical field of computer vision and virtual reality. Through a multi-viewpoint panoramic image set, real-time generation of a high-quality panoramic image from a known viewpoint to any unknown viewpoint is realized based on a spherical projection viewpoint transformation algorithm. Modeling generation methods for a road scene and a region scene are respectively designed according to scene feature differences, so that scene representation with complete coverage is constructed. According to the system, adjacent viewpoints are quickly positioned through a spatial index mechanism, geometric mapping and multi-viewpoint weighted fusion are performed by using the spherical projection algorithm, and a smooth and continuous free viewpoint roaming picture is rendered in a virtual reality environment. According to the method, on the premise of ensuring high fidelity of the scene, real-time rendering is realized on common consumption-level hardware, and the coherence and the degree of freedom of immersive interactive experience are improved.

The invention discloses a multi-robot process scheduling and cooperative control method based on artificial intelligence, and the method comprises the steps: carrying out the modeling of a process structure and robot resources, and generating a unified task information queue; environment perception and multi-source point cloud acquisition are carried out to generate multi-source point cloud data; carrying out Poisson curved surface reconstruction to obtain a three-dimensional reconstruction result and an occupied grid map; selecting a projection view angle and a grid resolution, and generating a multi-channel two-dimensional environment grid map; constructing an improved HarDNet network, and generating an environment semantic and risk map; performing task allocation and path planning to form a cooperative control plan; and performing cooperative execution, controlling the physical robot to execute tasks, and controlling the software robot to process businesses. Through combination of Poisson curved surface reconstruction and improved HarDNet network, unified scheduling and cooperative control of multi-robot tasks are realized.

The invention discloses an obstacle distance measurement method and system based on panoramic tiled display. Four-way or six-way fisheye cameras are installed on a vehicle, images around a vehicle body are collected, 3D rendering splicing is carried out based on calibration data (internal parameters / external parameters), and a 360-degree panoramic aerial view image is generated. A user clicks two points of a screen through a display terminal, a system converts screen coordinates into 3D normalized vertex coordinates through viewport inverse transformation, relative coordinate differences are calculated and then expanded into homogeneous coordinates, the homogeneous coordinates are converted into physical coordinate differences in combination with a pre-stored rendering transformation matrix inverse matrix (generated by model / projection / view transformation combination), and the 3D normalized vertex coordinates are converted into 3D normalized vertex coordinates. And finally calculating the actual physical distance between the two points. The system supports dynamic refreshing of a ranging result and renders a distance marking line on a panoramic image. According to the method, through multi-camera calibration, coordinate space mapping and a matrix transformation algorithm, interactive obstacle distance measurement in a vehicle-mounted panoramic display scene is realized, and the method is suitable for intelligent driving, vehicle auxiliary driving and other scenes.

The application provides a pinhole image splicing true value acquisition and splicing method based on a heterogeneous camera combination. The acquisition method first symmetrically installs two pinhole cameras on a support, and installs an fisheye camera between the two pinhole cameras. Based on a double spherical camera model, the three cameras are calibrated to obtain the internal parameters and external parameters of the three cameras. The left eye pinhole view, the right eye pinhole view and the middle eye fisheye view of the target scene are simultaneously acquired by using the three cameras. The middle eye fisheye view is subjected to equirectangular projection. According to the internal parameters and external parameters of the three cameras, the left eye pinhole view and the right eye pinhole view are projected to the equirectangular projection view plane respectively, multi-camera coordinate system one is completed, and the splicing true value is obtained. The application has the core advantages of the hardware structure design and the heterogeneous camera image registration algorithm, can stably acquire images without seams and time sequence consistency, and provides strong support for the true value reference of the precision verification and model training of subsequent panoramic video splicing.

One or more applications access a first set of database tables via a first projection view mapping from a first name to a second name. At a given point in time, the one or more applications may detect an indication of a modification operation targeting the first set of database tables. In response to detecting the indication of the modification operation, the one or more applications may create a second set of database tables for storing results of the modification operation. In response to detecting a completion of the modification operation, the first projection view may be remapped from the first name to a third name, where the third name is associated with the second set of database tables. Then, the one or more applications may access the second set of database tables via the first projection view mapping from the first name to the third name.

The application discloses a laser radar SLAM method based on a multi-modal structure semantic feature, and comprises the following steps: acquiring laser radar point cloud data; performing cylindrical point cloud segmentation on the laser radar point cloud data based on an ERFNet method; performing ground point cloud segmentation based on a BEV projection view method; performing planar point cloud segmentation based on a region growing method; fitting the cylindrical point cloud into a cylindrical model, and fitting the ground point cloud and the planar point cloud into planar models respectively; acquiring a matching relationship between different cylindrical models, and acquiring a matching relationship between different planar models, and constructing a structure semantic map; storing the structure semantic map based on a voxel plane; and updating the structure semantic map based on a model matching and map incremental updating algorithm. The application effectively improves the laser radar SLAM positioning precision, and proposes an intuitive multi-modal structure semantic map, so that the construction and real-time updating of three kinds of structure semantic model maps can be realized.

The utility model discloses a splicing type three-view building structure which comprises a display frame and a plurality of display square blocks, and a clamping block structure and a clamping groove structure are arranged on the two adjacent side edges of the display frame respectively. Crisscrossed dividing strips are arranged in the display frame and divide the interior of the display frame into N rows and N columns of square frame units, and N is larger than or equal to 3; the display square blocks are detachably placed in the square frame units; the clamping block structure comprises a plurality of clamping blocks, the clamping groove structure comprises a clamping groove strip arranged on one side of the display frame, and a plurality of clamping grooves in one-to-one correspondence with the clamping blocks of the clamping block structure are formed in the clamping groove strip in the length direction of the clamping groove strip; according to the utility model, manifestation forms of an object under different visual angles can be effectively displayed, so that children can simultaneously observe three-dimensional figures and projection views of three views, the children can be helped to establish conversion thinking between a two-dimensional plane and a three-dimensional space, and the development of early-stage space intelligence of the children is facilitated.

A method include obtaining original projection data of an object. The original projection data may be acquired by scanning the object using a radiation device from one or more projection views. The original projection data may include truncation projection data that is acquired from at least one of the one or more projection views. In each of the at least one of the one or more projection views a truncation region of the object may be located outside an FOV of the radiation device. The method may include obtaining a target S-shaped extrapolation function based on the original projection data. The method may include determining, based on the target S-shaped extrapolation function and the original projection data, extended projection data corresponding to the at least one truncation region. The method may include generating an image of the object based on the original projection data and the extended projection data.

1. The name of this design product: the main body of the analyzer. 2. Purpose of this design: The overall product is used to determine the composition of a sample, and this partial design is used as the main body of the analyzer. 3. The key design feature of this product is its shape. 4. The image or photograph that best illustrates the key design points: Design 1 3D view. 5. Designate Design 1 as the base design. 6. Other situations that need explanation Other explanations: The parts shown by dotted lines in the orthographic projection view and the stereoscopic view are the parts that do not require protection, and the parts shown by solid lines are the parts that require protection.

Methods and apparatus are disclosed herein to synthesize six degree-of-freedom views from sparce RGB-depth inputs. An example apparatus includes at least one memory, instructions in the apparatus, and processor circuitry to execute the instructions to reproject a first red-green-blue (RGB) depth view and a second RGB depth view to a target camera position to obtain a first reprojected view and a second reprojected view, combine the first reprojected view and the second reprojected view into a blended view data based on a first weight map and a second weight map, the blended view including missing RGB depth information due to at least one of an occlusion or a disocclusion, and generate a six degree-of-freedom synthesized view of the blended view data, the synthesized view including the missing RGB depth information.

Design 1 is a vertical beam of a shelf; Figs. 1.1 to 1.7 show six orthographic projection views and a perspective view of Design 1; Reference view 1.8 shows Design 1 in a status of use and forms no part of the claimed design.Design 2 is a horizontal beam of a shelf; Figs. 2.1 to 2.7 show six orthographic projection views and a perspective view of Design 2; Reference view 2.8 shows Design 2 in a status of use and forms no part of the claimed design.Design 3 is a horizontal beam of a storage rack; Figs. 3.1 to 3.8 show six orthographic projection views, one sectional view, and one perspective view of Design 3; Reference view 3.9 shows Design 3 in a status of use and forms no part of the claimed design.The products are made of metal; The double curved lines shown in the reproductions indicate the indeterminate length of the designs.

The invention discloses a GIS partial discharge diagnosis method based on a deep multi-view convolutional neural network, and the method comprises the steps: obtaining an ultrahigh frequency signal of GIS partial discharge, and extracting a physical parameter of the ultrahigh frequency signal to obtain a three-dimensional statistical graph containing a power frequency phase, a discharge capacity and the number of discharge times; inputting the three-dimensional statistical atlas into a partial discharge detection model; generating a plurality of orthogonal projection views from the three-dimensional statistical graph through coordinate transformation by using a multi-view projection layer; performing feature extraction on the plurality of orthogonal projection views by using a feature extraction layer to obtain feature maps of different scales; fusing the feature maps of different scales by using a feature fusion layer to obtain fused high-dimensional features; processing the high-dimensional features by using a full-connection classification layer to obtain a partial discharge type diagnosis result; according to the method, the calculation efficiency and the processing speed of fault diagnosis can be remarkably improved, and relatively high recognition accuracy and robustness can be kept under complex working conditions.

The present application belongs to the technical field of medical image processing, and particularly relates to a neural attenuation field method for iterative diffusion refinement of multi-source static CT reconstruction. In view of the reconstruction ill-conditioning problem caused by insufficient projection data under ultra-sparse view angle, the method first trains a neural attenuation field (NAF) model based on initial sparse projection; then adopts an angle prior guided strategy (APGPS) to adaptively determine a new projection view angle with the largest amount of information, and synthesizes the projection of the view angle by the NAF; then refines the synthesized projection by a double-branch conditional diffusion model (DRPR), and simultaneously predicts the structural residual and random noise; adds the refined projection as a pseudo-label to the training set, and iteratively updates the NAF model. The present application gradually expands the effective view angle coverage range through the iterative diffusion refinement mechanism, significantly suppresses the stripe artifact and restores the high-frequency details under the condition of no external labeled data, and breaks through the performance bottleneck of ultra-sparse view angle reconstruction.

1.1-1.6 represent six-sided orthographic projection views of an electrically heated air freshener in a combined state; 2.1-2.6 represent six-sided orthographic projection views of one component showing a main body of the electrically heated air freshener in the combination state; and 3.1-3.6 represent six-sided orthographic projection views of the other component showing a bottle for receiving the air freshener of the electrically heated air freshener in the combination state; 1.7-1.8, 2.7-2.8 and 3.7-3.8 respectively represent perspective views of the above designs; 1.9 represents a variation state view of the electrically heated air freshener in the combined state.

X-ray sources emit X-rays into an examination region along different projection views. An X-ray detector array detects the X-rays emitted by the X-ray sources after passing through the examination region. X-ray imaging data are acquired by cycling through the X-ray sources with each step of the cycle including: switching an active X ray source on to emit X-rays and the other X ray sources off to not emit X rays, and acquiring X ray imaging data along the projection view corresponding to the active X-ray source that is switched on. Tomosynthesis image reconstruction is performed on the X ray imaging data to generate at least one volumetric image. A time sequence of spatially aligned images is produced from the time sequence of volumetric images. A perfusion image and / or a ventilation image is generated based on voxel intensity variation over time of the time sequence of spatially aligned images.

Reproductions 1.1 to 1.8 are six orthographic projection views and two perspectives showing a dust box for cleaning appliance, reproductions 1.9 to 1.16 are reference views for representing a physical appearance of the dust box for cleaning appliance; the reference views are provided only for illustrative purpose and thus form no part of the claimed design.

The application discloses a cone beam CT geometry correction method suitable for irregular tracks, adopts a new optimization constraint iteration method based on a back projection model to solve geometry parameters under each projection view, and most of existing correction methods all assume that the internal geometry relation of a reference point in a phantom is accurate and known, and rarely deeply study whether the phantom precision is suitable for the correction model. Through the disclosure of the application scheme, the correction method is a high-noise robust correction method, and high-precision correction can be realized even when the resolution of a system to be calibrated is twice that of a phantom measurement system.

The application relates to the field of projection display technology, in particular to a windshield baking and bending mold. In the windshield baking and bending process, the baking and bending mold not only makes the whole windshield meet the bending requirement through an outer frame, but also is provided with a control block corresponding to the receiving and display light projection position of the windshield in the outer frame; the control block has a control surface corresponding to the projected area surface type of the windshield; the baking surface type fluctuation of the projected area of the windshield under the action of the self weight is reduced under the premise of controllable cost. The application can improve the stability of windshield processing, at least ensure that the surface type of the projection area meets the projection requirement of the HUD display device, and improve the projection viewing experience.

1. The name of the design product: integrated house (1). 2. The use of the design product: assembled integrated house for living, residence or office, etc. 3. The design points of the design product: the combination of shape and pattern. 4. The picture or photo that best indicates the design points: front view. 5. The six orthogonal projection views of the design have clearly expressed all the design features of the product, and the perspective view is omitted.

A projection system includes a control module, a projection tube, an aiming driver, an observation device and an observation driver. The control module is configured to issue a first control command and a second control command. The aiming driver is electrically connected to the projection tube and configured to control, in response to the first control command, a projection viewing-line of the projection tube to be aligned with a calibration point. The observation driver is electrically connected to the observation device and configured to control, in response to the second control command, an observation viewing-line of the observation device to be aligned with the calibration point. The projection tube and the observation device are controlled asynchronously.

The application discloses a kind of based on projection numerical control processing simulation method and medium, the method includes the following steps: the three-dimensional model of the workpiece to be processed is imported;Obtain the projection view of three-dimensional model, projection view includes front view, rear view, left view, right view, top view and overhead view;Surface machining feature information and depth cross-section machining feature information of the workpiece to be processed are obtained based on projection view;According to surface machining feature information and depth cross-section machining feature information, determine the region to be processed;According to depth cross-section machining feature information and the region to be processed, generate tool contact point trajectory;Based on tool contact point trajectory, generate numerical control processing trajectory, simulate.The application determines the region to be processed based on the projection six views of three-dimensional model, generates tool machining trajectory, can obtain higher precision processing area;And from three-dimensional model directly generate machining trajectory, without converting three-dimensional parts into two-dimensional drawing again processing trajectory design, greatly improve the efficiency of workpiece processing.

Reproductions 1.1 to 1.6 are six orthographic projection views of a first suction opener for small objects or models, reproduction 1.7 is a perspective view thereof; reproductions 2.1 to 2.6 are six orthographic projection views of a second suction opener for small objects or models, and reproduction 2.7 is a perspective view thereof.

The application relates to the technical field of artificial intelligence agent runtime security and audit governance, and discloses an execution permission verification and audit governance method and system for agent tool calling. The system receives a pre-execution stage standard behavior event before an agent executes a tool call, accesses data, publishes a scene or hands over a task, queries a permission policy library and a risk rule library based on operation identification, behavior subject information, permission context and risk marking, and outputs a disposal instruction; when permission is allowed or approved, an execution permission identification bound with the operation identification is generated, and a permission validity period, a permission range, an operation identification, a behavior subject, a scene layout version and a permission hash are verified by an external side effect pre-checking module before actual execution. The system also associates and saves the disposal instruction, the permission verification result and an actual execution result, and generates a desensitization projection view, so that the risk of overstepping the authority and exposing sensitive data is reduced.

A method is provided for determining optimized projection views for sharp object edge mapping of an object to be inspected using robotic computed tomography. The following steps are performed: determining a projection view in 3D Radon space, performing a collision check, and determining a consensus set as a subset of the points in 3D Radon space, where each point in 3D Radon space has a predefined weight. If the sum of the weights of the currently determined consensus set is higher than a value of one of the best current consensus sets, the currently determined consensus set is included in the best consensus sets, and the consensus set with the lowest value is removed.Upon reaching the termination criterion, the iterations are stopped, and the highest-scoring projection views are selected. A local check is then performed to determine the remaining pixel intensity in the resulting X-ray projection images from the projection views of the best consensus sets. The consensus set with the highest total score is selected and saved as the projection view to be used.

This invention discloses a direct-viewing near-eye display system and design method based on a transparent display screen with micro-nano luminescent pixels, belonging to the fields of augmented reality and near-eye display technology. Its core lies in: on the light-emitting side of the micro-nano luminescent pixels of the transparent display screen, a macro-pixel unit array is arranged, and a preset phase distribution is programmed for each area, directly modulating the display light to enter the human eye, eliminating traditional relay optics and achieving ultra-lightweight display. The system includes: a transparent display substrate, which also serves as a lens; a micro-nano luminescent pixel array located on the substrate, with the light-emitting side facing the human eye, forming a two-dimensional periodic macro-pixel unit array with a non-uniform density arrangement; and optical control elements disposed on the light-emitting side near the human eye, applying a preset phase distribution to each macro-pixel to achieve light collimation, deflection, and focusing, forming a Maxwell projection view. The remaining areas are transparent, used for observing real objects. Because the virtual image is directly focused into the eye, the imaging depth and physical distance are decoupled, eliminating convergence-accommodation conflict in principle. This invention has advantages such as high luminous efficiency, low power consumption, and lightweight design.

This invention relates to the field of computer-aided design technology, and discloses a method, apparatus, device, and storage medium for determining two-dimensional views of multiple parts. The method includes: obtaining corresponding blank entities and multiple part entities based on multiple projection entities; performing view projection on the multiple part entities to obtain corresponding multi-part entity projection views; determining position information and dimension information based on the multiple part entities and blank entities; and drawing the outline of the blank entity in the multi-part entity projection view using the position information and the dimension information. By performing view projection on the multiple part entities in the above manner, and then drawing the outline of the blank view in the multi-part entity projection view based on the position information and dimension information, compared with the prior art of obtaining part views by simultaneously projecting three-dimensional part entities and blank lines, the efficiency and accuracy of determining two-dimensional views of multiple parts can be effectively improved.

Provides a hat that exhibits improved breathability. [Solution] A hat X is provided, comprising a crown 1 and an air intake portion 2 that forms a pocket a on the outer surface of the crown 1, wherein the crown 1 has a ventilation portion 11 that connects the inside and outside of the crown 1, and the air intake portion 2 is provided such that the opening a1 of the pocket a faces forward, the internal space a2 of the pocket a is aligned substantially parallel to the front-rear direction, and covers at least a part of the ventilation portion 11, and the air intake portion 2 is provided on the outer edge of the crown 1 in a front projection view.

1. The name of the design product: bath ball (3). 2. The use of the design product: for bathing bath ball. 3. The design points of the design product: in shape. 4. The picture or photo that can best show the design points: perspective view 6. 5. Other front projection views have been embodied in the respective perspective views, and other front projection views are omitted.