210 results about "Spatial direction" patented technology

The invention discloses an ISAC angular domain energy imaging method and an ISAC angular domain energy imaging system based on actual measurement channel response calibration ray tracing, and mainly solves the problem that an imaging model faces serious channel mismatch and cross-scene generalization limitation due to the fact that simulation data lacks physical consistency constraint in the prior art. According to the scheme, the method comprises the steps that geometric features are extracted from public map data or laser point cloud data, and an ISAC scene environment geometric model is constructed; acquiring an actual measurement channel impulse response containing multipath information by using the communication and sensing integrated equipment; performing path matching on the actual measurement channel impulse response and the ray tracing simulation channel response, inverting environment propagation parameters and correcting scattering parameters; and inputting the calibrated parameters into a ray tracking model, counting energy contribution in each space direction under the condition of a single antenna, obtaining angular domain energy distribution and confidence coefficient, and mapping the angular domain energy distribution and confidence coefficient to a spherical grid to form an imaging result. According to the method, physical mismatch between simulation and real environments is effectively eliminated, low-cost and high-precision angular domain energy imaging and confidence evaluation are realized under the condition of a single antenna, and the method can be used for imaging reconstruction of an ISAC system, environmental perception and generation of a neural network training data set.

A quad-polarized antenna system according to the present invention comprises a plurality of RF chains, and an antenna array in which rows of first dual-polarized antenna units and rows of second dual-polarized antenna units are alternately arranged in a horizontal direction. The first dual-polarized antenna unit and the second dual-polarized antenna unit have different dual polarization characteristics from each other. The rows of first dual-polarized antenna units are used to form at least one first beam and at least one second beam which have + / −45° polarizations, and the rows of second dual-polarized antenna units are used to form at least one third beam and at least one fourth beam which have 0° / 90° polarizations. The first beam, the second beam, the third beam, and the fourth beam are formed towards different spatial directions, and the first and third beams correspond to main lobes, and the second and fourth beams correspond to grating lobes.

An optical system (100) for a projection exposure system (1), comprising an optical element (104, 106, 108), and a mount (132) which supports the optical element (104, 106, 108), wherein the mount (132) has a displacement device (138) which is configured to displace the optical element (104, 106, 108) linearly along exactly one spatial direction (152, 154, 156) in the event of a heat-induced deformation of the mount (132).

The invention belongs to the technical field of multi-modal vehicle obstacle avoidance, and relates to a multi-modal vehicle 4D dynamic obstacle avoidance method, which comprises the following steps: taking an air-ground amphibious vehicle, namely a multi-modal vehicle as a center, constructing a plurality of concentrically increasing sphere safety domains which respectively correspond to a collision avoidance area, an early warning area and a trajectory prediction area, the space-time position and the movement track of a dynamic obstacle are monitored and predicted in real time, a three-dimensional space is divided into eight quadrants according to the flight direction based on the current course of a multi-mode vehicle, different sensing priorities and obstacle avoidance strategies are given to the front quadrant, the side quadrant and the rear quadrant, and particularly, the front quadrant, the side quadrant and the rear quadrant have different sensing priorities and obstacle avoidance strategies. Dynamically adjusting the risk level and the obstacle avoidance response according to the relative speed and the movement trend of the obstacle, and carrying out hierarchical and regional risk perception and management on the dynamic obstacle in different space directions and distance ranges; the method is scientific and reliable in principle, and is suitable for a multi-body cooperative obstacle avoidance scene in an urban low-altitude traffic environment.

The invention discloses a multi-modal speech enhancement method and device based on a deep learning model, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining the head posture data, binaural audio signals and visual context information of a user in a virtual reality environment, coding the binaural audio signals into three-dimensional space acoustic features, and carrying out the coding of the three-dimensional space acoustic features; and extracting virtual sound source position features and lip motion features from the visual context information, inputting the features into an immersive fusion enhancement network, selectively enhancing or inhibiting acoustic features from different spatial directions, generating an enhanced audio stream, and outputting the enhanced audio stream through a binaural rendering engine. According to the method, the technical problems that in the prior art, due to the fact that multi-modal prior information cannot be effectively fused, voice enhancement lacks spatial selectivity, and an interference sound source irrelevant to vision is difficult to restrain are solved, and head posture dynamic attention and lip motion cross-modal constraint are fused; the technical effects of natural binding of auditory attention and a visual focus and effective suppression of an interference sound source are achieved.

A device (14) and a method for measuring wafers (10), comprising an optical coherence tomograph (22) configured to generate a measuring light beam (24) and direct it onto the wafer (10) via an optical system (38), a scanning device (26) configured to deflect the measuring light beam (24) in two spatial directions, a control unit (36) configured to control the scanning device (26) so that the measuring light beam (24) successively scans the surface of the wafer (10) at several measuring points, and an evaluation unit (57) configured to calculate distance values ​​and / or thickness values ​​from interference signals provided by the optical coherence tomograph (22), and a camera (80) configured to capture an image of the surface of the wafer (10), wherein the camera (80) images camera light (83).which has passed through the optical system (38) at least partially from the wafer (10).

The application provides a virtual-reality-based pronunciation device space direction parameter acquisition method, which only needs to perform parameter analysis on a reference pronunciation direction set, does not need to process remaining data in reference sound source data, can reduce the sound source data processing workload, can guarantee the accuracy of the sound source data to be processed, and reduces the data processing workload on the premise of reducing the sound source data to be processed. Moreover, since only the sound source data of the reference pronunciation direction set needs to be processed, the parameter recognition accuracy of the recognized pronunciation direction can be improved, so that the pronunciation direction parameter can be more accurate.

Method for imaging at least one beam (1) of electromagnetic radiation onto an object (2), wherein the beam (1) is deflected via its beam path (3) by an optical arrangement (4) by means of an optical scanner system (5) of the optical arrangement (4), comprising at least two optical scanners (6), at least twice with a predetermined and / or predefinable variable angle, each time changing the beam path (3) of the beam (1), and the beam (1) is not subject to refraction between each deflection, the beam (1) is also imaged onto the object (2) by means of an imaging optic (7) after its deflection, and a focus area (8) of the beam (1) is placed on and / or into the object (2) by means of the imaging optic (7).by means of the scanner system (5) at least one pivoting movement of a beam segment (9) of the beam (1) exiting the scanner system (5) is generated and the pivoting movement and / or at least a rotational movement (15) about the pivot point (10) causes a movement of the focus area (8) in at least one spatial direction (X, Y) perpendicular to the optical axis of the imaging optics (7), characterized in that during the pivoting movement the pivot point (10) of the beam segment (9) lies in the beam path (3) in the imaging optics (7) and beam shaping is imposed on the beam (1) via its beam path (3) by means of a beam shaping optic (16) by the optical arrangement (4), so that a focus area (8) extended in the direction of the beam path (3) is imaged on and / or in the object (2), which extends over at least a part of the dimension of the object (2) formed in the direction of the beam path (3).

A back-of-hand scanner (1) with an optical reading device (26) for reading a barcode on an object in a reading direction (7) in front of the back-of-hand scanner (1) and a trigger (27) which has an optical distance sensor (33) and triggers the reading device (26) when the distance sensor (33) detects a distance of the back-of-hand scanner (1) to the object below a limit distance, characterized in that the trigger (27) has an inertial sensor which includes acceleration and rotation rate sensors and detects translation in three independent spatial directions and rotation about the spatial directions and triggers the reading device (26) only when the inertial sensor detects a linear movement of the back-of-hand scanner (1) in the reading direction (7), wherein the trigger (27) blocks the triggering of the reading device (26) when the inertial sensor detects a rotation of the back-of-hand scanner (1).

This invention discloses an elastic wave reverse-time migration imaging method guided by stratigraphic dip angle for wavefield separation, relating to the field of seismic exploration technology. The method includes: acquiring an initial velocity model, source wavefield data, and receiver wavefield data for the target geological area; determining multiple stratigraphic dip angles reflecting the trend of subsurface structural changes based on the initial velocity model, and accordingly determining multiple target spatial directions; sequentially resampling, interpolating, and performing Hilbert transform on each wavefield data along each target direction to obtain source wavefield components and receiver wavefield components consistent with each direction; for any subsurface point to be imaged, determining the target spatial direction based on the stratigraphic dip angle corresponding to its vertical position, selecting wavefield components consistent with that direction for zero-delay cross-correlation calculation, thereby obtaining the reverse-time migration imaging result for that point. This method possesses good engineering feasibility and scalability, can be seamlessly embedded into existing pre-stack elastic migration workflows, and is convenient for the promotion and application of two-dimensional / three-dimensional and heterogeneous parallel computing.

This invention discloses a BIM-based method and system for calculating the quantity of outdoor pipe trenches, comprising: reading pipe parameters from the BIM model of the pipe components through a preset script program; automatically calculating the cross-sectional dimensions of the pipe trench based on the pipe diameter by calling a pre-constructed pipe trench cross-sectional parameter rule model; generating a pipe trench cross-sectional contour family by calling a parametric family library based on the pipe trench cross-sectional dimensions and the spatial direction coordinates of the pipe, and stretching the path along the pipe direction to generate a three-dimensional pipe trench model consistent with the pipe path; acquiring the volume data of the three-dimensional pipe trench model, calculating the excavation quantity of the pipe trench based on the volume data, and setting the excavation quantity into preset attribute fields to achieve dynamic statistics of the quantity. This invention can reduce the manual drawing and calculation process, improve the efficiency and accuracy of quantity calculation, realize the linkage update of model and data, and enhance the automatic adaptation capability to standard parameters.

A printhead (1) for parallel extrusion of printing medium, in particular for the production of conductive traces, is proposed, with a printing medium inlet (4) for the printing medium, which is fluidly connected to a plurality of output openings (5) of the printhead (1), for the simultaneous output of printing medium from the output openings (5). wherein the printhead (1) has a base unit (2) with the print medium input (4) and with a print medium output (6) as well as an output unit (3) detachably arranged on the base unit (2), wherein the output unit (3) has a connection area (7) via which it is connected or connectable to the base unit (2) in the area of ​​the pressure medium outlet (6), and has the plurality of output openings (5) for the pressure medium, which output openings (5) are fluidly connected to the pressure medium outlet (6) in a connected state of output unit (3) and base unit (2), which printhead (1) is characterized by the fact that in the area of ​​the pressure medium outlet (6) and the connection area (7) a flow path (F1) of the pressure medium is oriented in a first spatial direction (y) and in the area of ​​the output openings (5) the flow path (F2) of the pressure medium is oriented in a second spatial direction (z), wherein the second spatial direction (z) differs from the first spatial direction (y).

The invention relates to the technical field of computer technology and geological survey, in particular to a geological survey method based on big data. The method comprises the following steps: acquiring geological survey data of a sub-region, respectively preprocessing data of different data sources, and integrating basic feature vectors of the different data sources obtained by preprocessing; constructing a geologic structure map based on the fused feature vectors of the sub-regions obtained through integration, and introducing an edge weight construction mechanism combining a spatial adjacency relation and feature similarity to obtain edge weights among the sub-regions; and based on the edge weight between the sub-regions, introducing a directivity and edge weight joint modulation mechanism, carrying out node feature propagation, and predicting a geological structure category. The problems that a traditional geological survey method cannot accurately evaluate the credibility of each data source during fusion, neglects the correlation between the regions on the feature level, and is difficult to sense the composite action of the spatial direction difference between the regions and the edge connection strength on the propagation path are solved.

The invention belongs to the field of computer vision, and particularly relates to a multi-mode CNN-Transform fused image tampering detection method, which designs a CNN and Transform double-flow parallel feature extraction structure, effectively combines the advantage of CNN at capturing fine local features and the advantage of Transform in capturing long-distance dependency relationship and global semantic information, and improves the accuracy of image tampering detection. Local details and global tampering features in the image can be sensitively perceived at the same time, so that the detection capability and stability of image tampering are effectively improved. Processing the noise domain image filtered by the SRM through a CNN (Convolutional Neural Network), and capturing local texture and noise artifact features; meanwhile, a Transform branch processes an RGB spatial domain image, extracts rich global semantic information, and realizes deep interaction and advantage complementation of two types of modal information through a feature fusion module. The BAFM module provided by the invention can deeply mine feature information of different scales and different spatial directions, and more accurate spatial attention features are generated, so that the sensitivity and expression ability of the network to tampered regions are improved.

The invention relates to a ground treatment device which travels autonomously, having an electric drive for moving the ground treatment device, an obstacle detection device for detecting obstacles in the environment, a sensor device for detecting parameters of the ground being driven over by the ground treatment device, and an evaluation and control device for creating an environmental map from the detected obstacle data and parameters. In order to investigate further properties of the ground which can advantageously be taken into account when planning the driving strategy of the ground treatment device, it is proposed here that the evaluation and control device is set up to rotate the ground treatment device around a defined position point of the ground, the evaluation and control device is also set up to analyze the parameters detected by means of the sensor device when the ground treatment device is rotated or moved around the defined position point of the ground, and to determine the anisotropy of the parameters of the ground with reference to different spatial directions defined in the environment, and to record the anisotropy in the environmental map with the angular coordinates specified.

The application discloses a kind of based on neural network's image processing method and system, belong to image recognition technical field, wherein method includes based on initial tensor obtains second attention tensor and multiple first attention tensor, and based on multiple first attention tensor and second attention tensor obtains output tensor.Through setting two parallel network structures, respectively from the interrelation of feature map channel and the image space relation of each channel two dimensions carry out attention calculation, by down-sampling-attention relation calculation-up sampling design realizes to feature map each pixel attention modeling while reducing calculation amount;Finally, the attention information of channel and spatial direction is fused, obtains attention weight matrix, can carry out pixel-level channel and spatial attention modeling to feature map tensor, effectively improve the correlation extraction and modeling capability of convolutional neural network to image feature under large scale.

A configuration that indicates associations is transmitted from a network device of a communication network and received at a user equipment (UE). The associations are between sets of locations and spatial directions, such that each set of locations is associated with a respective one of the spatial directions. Each set of locations is a set of source locations of a respective reference signal. One of the references signals is transmitted from the network device and received at the UE, and multipath measurements associated with the source locations of the received one of the reference signals are performed at the UE.

The invention relates to a method and a device for determining a spatial pose of a non-spherical object (110) comprising at least one surface portion (111) having a known shape. The method comprises the following steps: a) radiating at least two light fields (100) onto the object (110) through a lens (70), each light field (100) having a known phase wavefront (104); the phase wavefronts (104) of the at least two light fields (100) differ from each other at least downstream of the lens (70); the curvature of the phase wavefront (104) at the exit window of the lens (70) is smaller in at least one spatial direction than the minimum curvature of the surface portion (111) of the object (110); for at least one of the light fields (100) there is at least one line located in the surface portion (111) along which at least two local surface normals (112) of the surface portion (111) are orthogonal to a phase wavefront (104) of the at least one of the light fields (100); b) detecting, by means of a detector (80), at least two reflected portions (102) of the light field (100) originating from the object (110); and c) determining the spatial pose of the non-spherical object (110) on the basis of the detected reflected portions (102) of the at least two said light field (100) in view of an expected backscatter signal generated by the surface portion (111) of known shape The method and device preferably enable the determination of the spatial pose of a cylindrical object, in particular glass fiber.

Method for detecting magnetic flux densities of a gradient coil (18) of a magnetic resonance device (1) within a patient tunnel (14) of the magnetic resonance device (1), comprising: determining (40) at least one position within the patient tunnel (14) at which the magnetic flux densities of a B0 magnetic field (13) of a B0 main magnet (12) of the magnetic resonance device (1) are below a predefined value in a first spatial direction and in a second spatial direction; arranging (41) at least one magnetic field strength sensor (35) at the at least one position, wherein the magnetic field strength sensor (35) is arranged and configured to detect the magnetic flux density at least in the first spatial direction and in the second spatial direction; and detecting (42) the magnetic flux densities by the at least one magnetic field strength sensor (35) in the first spatial direction and in the second spatial direction.

The invention relates to a control element (10) comprising a light source (12), a symbol layer (14) and a barrier layer (16), wherein the light source (12) is configured to illuminate the symbol layer (14); the symbol layer (14) has a first symbol (18) and a second symbol (20) for display on the control element (10), which can be illuminated by the light source (12); and wherein the barrier layer (16) is configured and arranged to allow first light rays from an illumination of the first symbol (18) to pass through in a first spatial direction (24) and to block them in a second spatial direction (26), and to allow second light rays from an illumination of the second symbol (20) to pass through in the second spatial direction (26) and to block them in the first spatial direction (24).

A piezoelectric composite energy harvester (100) is disclosed, comprising a piezoelectric element (102) and an elastic cover layer (110). The piezoelectric element comprises a first electrode (104), a second electrode (106), and a piezoelectric layer (108), the piezoelectric layer (108) comprising a cross-linked silicone elastomer matrix and piezoelectric particles dispersed within the cross-linked silicone elastomer matrix. The elastic cover layer (110) is disposed over the exterior of the second electrode (106). The piezoelectric composite energy harvester is configured to bend, expand, contract, rotate, tilt, and / or move in any spatial direction.

The invention relates to a ceiling suspension device (100) for lifting, transferring and lowering containers, comprising at least one crawler-type ceiling unit (10) and a ceiling structure (1) extending in at least two spatial directions (x, y), the ceiling structure comprising a plurality of profile units (1.1) extending in a first spatial direction, wherein the ceiling structure defines at least one structural regularity (1a) in at least one region, and wherein the crawler-type ceiling unit is configured to move in said first spatial direction (x) irrespective of instantaneous movement in said second spatial direction (y); and wherein the at least one ceiling unit comprises at least one elevator unit (50) wherein the at least one elevator unit comprises a gripping unit wherein the gripping unit (53) is configured to be connected and / or connected to at least one container (60). The invention further relates to a port comprising the suspended ceiling suspension device.

Household appliance (1) with at least one user interface (13) for input and output of information and a door (3) which has an edge (14, 14'), wherein the at least one user interface (13) extends partially or completely around the edge (14, 14'), wherein the at least one user interface (13) has a first surface (15) and a second surface (16), wherein the first surface (15) points in at least a first spatial direction and the second surface (16) points in at least a second spatial direction, and wherein the first and second spatial directions are different, and wherein, when the door (3) is closed, the first surface (15) of the at least one user interface (13) points upwards and the second surface (16) of the at least one user interface (13) points forwards, characterized in that the first surface (15) and the second surface (16) of the at least one user interface (13) each serve for input and output of information.

A communication method and apparatus, relating to the field of communication technology, can improve the reliability of communication between multiple terminal devices and network devices while minimizing resource overhead. The method includes: a network device sending first configuration information to each terminal device in a first set of terminal devices; the first configuration information includes DMRS port information and ZP-DMRS port information corresponding to the terminal device; sending second configuration information to each terminal device in a second set of terminal devices; the second configuration information includes DMRS port information corresponding to the terminal device; DMRS ports corresponding to terminal devices belonging to the same set of terminal devices are orthogonal, while at least two DMRS ports corresponding to different sets of terminal devices are non-orthogonal; the first set of terminal devices and the second set of terminal devices are determined based on a first spatial direction corresponding to the terminal device, and the first spatial direction corresponding to the terminal device is determined based on the energy of the spatial direction corresponding to the terminal device.

The invention relates to a cable unit designed for connecting a gradient coil, which is designed to generate a magnetic field gradient in one spatial direction, to a power amplifier, comprising at least one supply line, wherein the supply line comprises at least two cable elements in at least sections, wherein the at least two cable elements are electrically connected in parallel.

A method for planning an autonomous landing safe trajectory suitable for UAV inspection is proposed. After the UAV takes off, it acquires the current pose of the UAV, the point cloud map information of the surrounding environment, and the spatial orientation of the target point. By establishing the spatial transformation relationship between the camera coordinate system and the body coordinate system, the orientation vector relative to the ground target is calculated, and the target point is generated within the radar-sensible area. Based on the generated target point, an optimization function that minimizes the fourth derivative is constructed based on multiple high-order polynomials to generate a smooth trajectory with continuous position, velocity, and acceleration. The initial trajectory is transformed into a B-spline curve, and local trajectory segments with collision risk are optimized. A segmented penalty mechanism is designed based on the boundary threshold to guide the trajectory away from the boundary area. The time span is adjusted by calculating the trajectory derivative over-limit ratio, and a trajectory similarity cost function is introduced to ensure that the final optimized trajectory maintains the original geometric structure, thereby achieving trajectory control optimization and safe landing of the UAV.

The invention discloses a DAS signal conversion method and device based on a frequency wavenumber domain, and the method comprises the steps: obtaining an external excitation signal, and determining the phase difference information of Rayleigh scattering light in the excitation signal; demodulating the Rayleigh scattering light to obtain an electric signal, and obtaining a strain signal based on the electric signal; converting the strain signal into a frequency-wave number domain signal based on a first conversion formula, and performing denoising processing on the frequency-wave number domain signal to obtain a denoised strain signal; integrating the denoised strain signal along the spatial direction of the optical fiber based on a second conversion formula to obtain a relative displacement signal of each point along the direction of the optical fiber; obtaining a speed signal of the mass point based on the relative displacement signal, and obtaining an acceleration signal of the mass point based on the speed signal of the mass point; and forming a detector response based on the acceleration signal of the mass point, thereby improving the conversion efficiency and conversion quality of the DAS signal during time-space domain conversion.

The application belongs to the field of semantic segmentation, and provides a method and system for road scene semantic segmentation. The method comprises the following steps: based on an obtained road scene image, a ResNet101 network is used to obtain a first feature map; the first feature map is input into a global spatial position relationship module to calculate the inner product between a pixel and other pixel points in a spatial direction, so as to obtain a second feature map representing the relationship between the pixel and the pixel in the spatial direction; the second feature map is input into a global channel position relationship module to calculate the inner product between the pixel and other pixel points in a channel direction, so as to obtain a third feature map representing the relationship between the pixel and the pixel in the channel direction; the second feature map and the third feature map are fused to obtain a fourth feature map representing the long-distance dependence relationship between any two pixel points; and different categories of objects are predicted pixel by pixel according to the fourth feature map.

This invention discloses a mechanical superstructure design method and a mechanical superstructure designed using this method, belonging to the field of mechanical superstructure technology. The method includes establishing a finite element model of a multistable unit cell, which comprises a support frame and multiple sets of arc-shaped beam structures, each set consisting of two intersecting arc-shaped beams; setting initial geometric parameters for each set of arc-shaped beam structures; obtaining the jump load thresholds for the arc-shaped beam structures in each direction through finite element analysis; determining whether the jump load thresholds in each direction increase or decrease sequentially according to a preset spatial direction; if not, adjusting the corresponding geometric parameters and iteratively solving until the sequence requirement is met, forming a programmable multi-axis jump sequence; outputting the optimized final geometric parameters; and constructing multiple multistable unit cells based on the final geometric parameters and assembling them into an array structure along spatial directions. This invention provides a mechanical superstructure capable of independent deformation in multiple spatial directions and with a programmable jump sequence.