968 results about "Direction information" patented technology

The invention relates to the technical field of aviation communication, and discloses a true north calibration system and method for an aviation airborne laser communication equipment alignment system, and the method comprises the steps: synchronously collecting flight data through an IMU sensor and a double-antenna GPS at a preset sampling frequency, and outputting the real-time attitude angle information of an aircraft through an IMU; filtering and de-noising the synchronously acquired IMU and double-antenna GPS data, and carrying out noise suppression and deviation compensation on accelerometer and gyroscope data output by the IMU by adopting a Kalman filtering algorithm; based on the preprocessed sensor data, calculating an initial course angle by measuring relative position change and phase difference between antennas by using a double-antenna GPS (Global Positioning System); the calibration controller transmits the calculated true north direction information to an alignment system of the laser communication equipment in real time through a communication interface; and the attitude change and external environment disturbance of the aircraft are continuously monitored in the whole flight process. The method has the advantage that the alignment performance and the communication stability of the laser communication equipment are improved.

The invention provides a smart glasses AI voice interaction method, which comprises the following steps: simultaneously acquiring a user gesture image and a voice signal, acquiring a user historical interaction record, extracting key point coordinates and pointing direction information of the gesture image, acquiring hand distance information, generating a gesture motion path, simultaneously extracting frequency spectrum information and an intonation peak value of the voice signal, and generating a gesture motion path; forming a voice beat sequence; extracting a spatial semantic mode of the gesture and voice fusion data, and recognizing a core object of a user pointing instruction according to pointing coordinates of a gesture motion path and an intonation peak value of a voice beat sequence; after the core object pointing to the instruction is recognized, the user intention is determined in combination with the spatial semantic mode and the historical interaction record of the user; and the complete intention analysis result is output to the intelligent glasses display module to execute corresponding operation, and is fed back to the acquisition module to adjust the next capture parameters including the acquisition frequency and the recognition sensitivity, so that the response speed and the accuracy are improved.

The invention relates to the technical field of quality detection, in particular to a product quality control method and system based on machine vision, and the method comprises the following steps: obtaining product surface image data, calculating the gray gradient value of each pixel, extracting the gray gradient change rate, recording the gradient amplitude and direction information, and generating product surface gradient data. According to the method, through pixel-level gray scale gradient calculation, the product local feature expression ability is improved, multi-scale gradient change trend analysis is combined, the accurate recognition ability of a product defect area is improved, through texture direction angle calculation and vector field construction, the direction change anomaly detection reliability is enhanced, and the direction change anomaly detection accuracy is improved based on the combination of a direction deviation accumulated value and an abrupt change threshold value. Effective identification of a structure sudden change area is ensured, adjustment is carried out for curvature continuity abnormal points, defect boundary fitting precision is optimized, defect area internal gradient distribution and boundary feature comparative analysis are carried out, accurate classification of defect types is realized, and stability and adaptability of automatic product quality detection are ensured.

The invention belongs to the technical field of plastic part quality detection, and discloses a plastic part gold stamping mark accurate positioning method and system based on visual identification, and the method comprises the steps: collecting environment illumination parameters, equipment operation parameters and image data, and processing the image data to obtain regional image data; performing edge detection and fracture repair operation on the regional image data to obtain identification edge feature data; calculating to obtain a position deviation quantized value according to the identification edge feature data; based on the position deviation quantized value, in combination with preset inclination direction information, adjusting an image visual angle and correcting an angle deviation to obtain corrected image data; executing edge path tracking on the corrected image data to obtain path complete data; obtaining an identifier positioning result based on the path complete data in combination with the environment illumination parameters and the equipment operation parameters; and carrying out stability evaluation on the identification positioning result, and obtaining a final system identification stability result in combination with preset detection fluctuation characteristics.

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of intelligent auxiliary navigation, financial science and technology, medical health and the like, and discloses a navigation instruction generation and path updating method, device, equipment and medium. And fusing the obstacle category information and the environment depth information, generating spatial position information and direction information of the obstacle, generating a navigation instruction containing scene description information, obstacle orientation and distance description information and action suggestion information based on the spatial position information, the direction information and the environment image information, and updating a navigation path. According to the method, the obstacle category information and the environment depth information are fused, the obstacle recognition result with the semantic level is generated in combination with the environment image information, the navigation instruction containing the orientation and the distance is output, and the accuracy of obstacle perception and the practicability of navigation guidance can be improved.

The invention discloses a cosmic ray muon scattering imaging method, system and device, and relates to the field of ray imaging, and the method comprises the steps: obtaining the direction information of muons in a to-be-imaged region; calculating the most probable three-dimensional scattering track of the muons in the to-be-imaged region based on Bayesian inference and a particle multiple coulomb scattering distribution function according to the direction information of the muons in the to-be-imaged region; and reconstructing a scattering density image in the to-be-imaged region according to the direction information of the muon in the to-be-imaged region and the most probable three-dimensional scattering track. By calculating the most probable three-dimensional scattering track, the scattering points can be effectively reconstructed, the utilization efficiency of muon events is improved, the scattering points are prevented from being separated from the to-be-imaged area, the most probable three-dimensional scattering track better fits the real scattering track, and then the quality of the scattering density image is improved.

The invention relates to the technical field of image analysis, in particular to a motorcycle paint surface bubble sagging defect real-time detection system based on image analysis. The method comprises the following steps: firstly, acquiring a plurality of images of a to-be-detected paint surface under different illumination conditions by an image acquisition and processing unit, and preprocessing the images to obtain a preprocessed multi-illumination image; then, a three-dimensional information analysis unit analyzes and calculates a surface normal vector of each pixel point of the paint surface according to brightness information of the same pixel point in the preprocessed multi-illumination image, and generates a surface normal graph for representing three-dimensional direction information of the paint surface; then, a candidate area positioning unit analyzes a gradient field of the surface normal diagram and positions bubble and sagging defect candidate areas respectively; and finally, the feature extraction and classification unit extracts fusion features of the defect candidate region, and outputs a detection result obtained by recognition of the classifier through an output unit. According to the invention, the accuracy of paint bubble sagging defect detection can be improved.

The invention provides a single-microphone multi-array pickup method and system based on sound attenuation simulation, and the method comprises the steps: obtaining an original audio signal recorded by a single microphone and sound source direction information, and carrying out the sound attenuation simulation based on the sound source direction information and a preset virtual microphone array orientation parameter, calculating the included angle between the sound source direction and each virtual microphone; generating an audio intensity attenuation coefficient of each virtual microphone according to the included angle; and acting the audio intensity attenuation coefficient on the original audio signal to generate multi-channel audio data simulating the multi-microphone array. By adopting the method, the response difference of different microphone positions to the sound source can be simulated, the pickup result of the multi-microphone array is obtained, and high hardware cost and complex installation flow caused by actual deployment of a complex microphone array are avoided; and low-cost and high-diversity training data sources are provided for sound source localization, noise suppression and other models based on deep learning.

The invention provides an arbitrary point turn-back method under a train control system route architecture based on communication, which comprises the following steps of: when an arbitrary point turn-back command for a train is received and the train is determined to be in a stable stop state, generating a reverse route containing turnout opening information. And according to the reverse direction head code, generating a starting end signal machine list of which the route needs to be manually solved and sending the starting end signal machine list to an automatic train protection system. And an unlocking request command of the first signal machine in the starting end signal machine list is sent to the area controller, and the access in the starting end signal machine list is unlocked through interactive verification with a train automatic protection system. After it is confirmed that the route corresponding to the first signal machine is successfully unlocked, a turn-back instruction is sent to the automatic train protection system, and meanwhile the automatic train protection system is controlled to execute end changing operation. After the end changing operation is completed, new movement authorization is established to control the train to run, route unlocking in the original direction and passenger prompting are triggered when preset conditions are met, meanwhile, a turn-back mark of any point is updated and displayed on a human-computer interface, and turn-back of any point is achieved.

The present application relates to the field of federated learning and privacy protection, and particularly relates to a direction-aware local differential privacy federated learning method. The present application effectively solves the problem of serious decline in model performance caused by the destruction of gradient direction information by traditional differential privacy methods by adopting a direction-aware noise selection strategy based on an exponential mechanism locally on the client side. Specifically, the client first performs L2 norm clipping on the gradient to control the sensitivity, then generates K candidate noise vectors from a Gaussian distribution, evaluates the direction consistency of each candidate noise and the true gradient through cosine similarity, and selects the noise closest to the gradient direction in a probabilistic manner using the exponential mechanism, and finally adds the selected noise to the gradient to complete the perturbation. The method provides a strict epsilon-local differential privacy mathematical proof for the entire noise selection process, reduces the calculation time by about 80% compared to existing direction-aware methods, and achieves a triple balance of privacy protection, model utility and computational efficiency.

The invention discloses a two-dimensional construction drawing exporting method and system based on a three-dimensional BIM model, and the method comprises the steps: analyzing the geometric information and attribute information of a component from obtained three-dimensional BIM model data, and the geometric information comprises pipeline geometric information, equipment geometric information and structure geometric information; on the basis of the analyzed geometric information, automatically identifying a labeling object; carrying out labeling processing on the identified labeling object, calculating an optimal angle and a character avoidance position of a labeling lead in a labeling process, and generating optimized labeling information; on the basis of the analyzed attribute information, automatic layer mapping is conducted on the component, and standardized layer information is generated; on the basis of the analyzed pipeline geometric information, performing axonometric projection transformation on topological relation data obtained through pipeline topological relation analysis, and generating axonometric map information including a pipeline space connection relation and flow direction information; and integrating the labeling information, the layer information and the axonometric drawing information, and outputting a standardized two-dimensional construction drawing.

The invention discloses a brain tumor MRI image semantic segmentation method, and belongs to the technical field of medical image processing and computer-aided diagnosis. According to the invention, the problem of low segmentation precision obtained based on the existing brain tumor MRI image semantic segmentation technology is solved. The invention provides a semantic segmentation model which combines a U-Net framework with an LWD module, an MFDF module and various filters, the LWD module can keep information as much as possible in a down-sampling process, the MFDF module extracts operators by constructing new directional gradient features, and combines the operators in different directions by using dual-channel filtering, so that the semantic segmentation of the U-Net framework is realized. And the constructed multi-directional filter can respectively extract low-frequency and high-frequency characteristic direction information. The MFDF module transmits detail features from the coding module to the corresponding decoding module, so that spatial information including feature boundaries and textures is recovered, the precision of model segmentation is improved, and the method has good adaptability to the randomness of brain tumor shapes, sizes and boundaries. The method can be applied to brain tumor MRI image segmentation.

The invention discloses a visual sag detection method, system and device. The method comprises the following steps: S1, pre-calibrating external parameters of a camera and a laser radar, and obtaining internal parameters and relative external parameters of the camera and the laser radar; s2, synchronously acquiring image data and three-dimensional point cloud data of the to-be-detected power transmission conductor respectively; s3, performing fitting by using the wire local three-dimensional point cloud to extract a wire part point cloud, and determining a vertical plane where the wire is located by combining gravity direction information and external parameters; s4, segmenting image pixels of the leads in the image acquired by the camera, and determining pixel coordinates of the leads in a camera coordinate system; and S5, projecting the segmented conductor image pixels to a vertical plane, expanding the conductor image pixels into a three-dimensional space point cloud under a camera coordinate system, fitting a catenary of the power transmission conductor line based on the three-dimensional space point cloud, and calculating a sag value of the conductor. According to the invention, the sensing precision of the conductor state in the sag detection technology can be improved, and high-precision sag detection can be continuously carried out after a single system is installed.

The invention discloses an unmanned aerial vehicle autonomous positioning navigation method and device based on a cross-view azimuth regression network. The method comprises the steps of 1, constructing a cross-view-angle data set which is oriented to a cross-view-angle scene and contains azimuth information, 2, constructing an unmanned aerial vehicle position heading regression model oriented to the cross-view-angle scene, and 3, training and verifying the unmanned aerial vehicle position heading regression model oriented to the cross-view-angle scene by using a training set and a verification set. And 4, carrying out regression performance test on the unmanned aerial vehicle position heading regression model facing the cross-view scene by using the test set. The method only depends on a visual image, can reduce the dependence degree of unmanned aerial vehicle positioning navigation on a sensor, improves the survivability in an extreme environment, uses a cross-visual-angle remote sensing map to construct simulation training data and generate the absolute orientation regression capability, can eliminate the drift problem, and also has natural advantages in the aspect of migration from simulation to a real scene. The model is light in weight and is more suitable for unmanned aerial vehicle continuous positioning and directional navigation scenes with limited computing power resources.

The invention relates to the field of wind power generation, and discloses a wind power generation control system and method. The data acquisition module is used for acquiring front three-dimensional wind field information detected by the laser radar, near-field wind speed and direction information measured by the ultrasonic sensor, regional wind regime data issued by the meteorological satellite and operation parameters of the wind generating set. According to the invention, through multi-source information fusion, wind regime accurate perception is realized, power generation stability is improved from the source, through multi-source fusion of the laser radar, the ultrasonic sensor and meteorological satellite data, a three-dimensional wind regime perception system is constructed, wind speed and wind direction changes can be predicted in advance, wind regime measurement errors are reduced, and power generation efficiency is improved. The problem of power generation fluctuation caused by wind energy intermittency is solved from the source, and accurate data support is provided for dynamic adjustment of the pitch angle.

The invention discloses a multi-stage text-to-three-dimensional model generation method, and particularly relates to the technical field of artificial intelligence, comprising three stages of initialization, optimization and clone segmentation; the method comprises the following steps: firstly, generating an image and depth through a multi-view diffusion model, and guiding to construct an initial three-dimensional Gaussian point cloud; secondly, multi-step diffusion inversion is executed under micro rendering, and the point cloud is optimized in combination with consistency loss and structure regularization; and finally, according to a significance score and a detail hierarchy mechanism, in combination with curvature and principal direction information, executing local clone splitting, and improving structural detail expression and density regulation and control capabilities. The method has the advantages of high semantic fidelity, high structural consistency and the like, and is suitable for scenes of text-driven three-dimensional modeling, virtual reality, point cloud reconstruction and the like.

Provided are an apparatus and method for operation of an apparatus in a wireless communication system. The apparatus may be a UE, a chipset, or a base station. The UE receives an FD configuration message notifying a full duplex symbol from among a plurality of symbols, receives a cell-specific common TDD configuration message, and receives a UE-specific dedicated TDD configuration message. Here, only for the full duplex symbol from among the plurality of symbols, second direction information notified by the dedicated TDD configuration message overrides first direction information notified by the common TDD configuration message.

Embodiments of the present application disclose an automatic parking control method and device and a computing device. The method comprises: acquiring first position information of a position where a vehicle is located, a vehicle inclination angle, second position information of a target empty parking space, and direction information; judging whether the vehicle is inclined relative to the target empty parking space based on the vehicle inclination angle and the direction information of the target empty parking space; if the vehicle is not inclined relative to the target empty parking space, determining a lateral movement distance and a lateral movement direction of the vehicle according to the first position information and the second position information; and controlling the vehicle to move laterally into the target empty parking space based on the lateral movement distance and the lateral movement direction, which helps to improve the efficiency of automatic parking.

A computer-assisted medical system includes a user device wearable by an operator. The user device includes a display device configured to present imagery overlaid in an environment of a manipulator assembly, and a sensor configured to detect one or more landmarks in the environment. The medical system includes a controller configured to execute instructions to perform operations The operations include receiving, from the sensor, position or orientation information for the one or more landmarks in the environment, and directing a manual movement of a portion of the manipulator assembly by causing the display device to present the imagery overlaid in the environment based on the received position or orientation information.

The invention discloses a multi-type agent trajectory prediction method and system fused with a multi-graph convolutional network, and the method comprises the following steps: S1, obtaining scene information and trajectory information of multiple types of traffic participants, and generating a data set; s2, generating scene feature maps according to scene information in the data set, and weighting and fusing the scene feature maps through the attention score to generate global scene information; s3, respectively extracting position information, speed information and direction information of various types of traffic participants in the data set, and constructing a multi-influence graph; s4, jointly embedding the processed multi-influence graph and global scene information into a space-time graph convolutional neural network, and performing coding processing through the space-time graph convolutional neural network to generate dynamic space-time features; and S5, decoding the dynamic spatio-temporal characteristics through the spatio-temporal transformation trajectory generation network to obtain future moment trajectory points of the traffic participants. According to the method, the dynamic interaction relationship among the multiple types of traffic participants can be more accurately captured in trajectory prediction of the multiple types of traffic participants, and the prediction precision and the real-time response capability in a complex traffic environment are improved.

The invention discloses a bolt surface defect detection method and system based on computer vision. The method comprises the steps that 360-degree image collection and preprocessing are conducted on a bolt through an annularly-arranged multi-camera array; fourier transform is adopted to analyze thread periodic texture features, and an adaptive filter is designed to enhance the image; multi-scale morphological operation is adopted, and morphological gradient features are extracted by using structural elements for crack, pit and scratch type defects; a self-adaptive threshold segmentation method based on local statistical characteristics is combined with thread direction information to dynamically adjust a segmentation threshold, and defect candidate areas are segmented; and extracting geometric and textural features of the candidate region, and identifying the type, position and severity of the defect by adopting a random forest classifier. According to the invention, high-precision automatic detection and classification of various defects on the surface of the bolt are realized.

The invention discloses a neural radiation field three-dimensional reconstruction method and system based on a variational auto-encoder, and the method comprises the steps: constructing a variational auto-encoder, carrying out the feature coding of scene data in a training set, and generating a low-dimensional potential vector representing the global information of a scene; carrying out high-dimensional coding on the three-dimensional space coordinates and view angle direction information of the sampling points, and splicing a high-dimensional coding result with the low-dimensional potential vector to form an enhanced input vector; constructing an adaptive sampling algorithm, and adjusting a sampling strategy of a neural radiation field based on importance distribution of a scene region to realize high-precision sampling of a key region; constructing a neural radiation field optimization model fusing a variational auto-encoder and adaptive sampling, training the neural radiation field optimization model by using the enhanced input vector, and realizing three-dimensional reconstruction of a target scene through optimized neural radiation field parameters; according to the method, the calculation cost can be reduced, the sampling efficiency is improved, and the reconstruction integrity under sparse data is enhanced.

The invention provides a remote sensing image rotating target detection method based on dual-path feature enhancement, and relates to the technical field of computer vision and remote sensing image processing. The method comprises the following steps: constructing a dual-path feature enhanced remote sensing image rotating target detection network comprising a texture enhancement path and a direction modeling path; wherein in the texture enhancement path, a self-adaptive wavelet reconstruction module is adopted to enhance texture details and edge features in the input feature map; in the direction modeling path, performing spatial alignment and direction consistency modeling on the input feature map by adopting a multi-scale angle guide deformable encoder, and extracting features containing structure and direction information; fusing the features output by the two paths; constructing a joint loss function, and performing end-to-end training on the network; and detecting and positioning a rotating target in the remote sensing image by using the trained network. By adopting the method, the detection precision and robustness of multi-direction, multi-scale, densely distributed and small-size targets in the remote sensing image can be effectively improved.

Provided is an information processing apparatus including a processor; and a memory that is connected to or incorporated in the processor, in which the processor acquires imaging device position information regarding positions of a plurality of imaging devices and imaging direction information regarding an imaging direction of each of the plurality of imaging devices, derives virtual viewpoint range specification information capable of specifying a virtual viewpoint range in which a virtual viewpoint image is capable of being generated on the basis of a plurality of captured images obtained by being captured by the plurality of imaging devices, on the basis of the imaging device position information and the imaging direction information, and outputs information regarding the virtual viewpoint range specification information.

The invention discloses a water conservancy management system and method based on digital twinning, and relates to the technical field of water conservancy management, and the method comprises the steps: collecting a water flow sound wave signal through vibration optical fiber sensors disposed in a water conservancy facility and a water channel, and converting the water flow sound wave signal into a digital pulse signal carrying flow velocity, impact force and flow direction information; constructing a water conservancy topology network based on a water conservancy project drawing; geographic units are divided by expanding a preset distance along the boundary of the water channel, and a geological memory capsule is created for each geographic unit; when a scheduling instruction is received, retrieving a target geological memory capsule in an influence range of the scheduling instruction, and calling an evolution operator of the target geological memory capsule to deduce geological responses of different time scales in parallel; and performing spatial overlay analysis on the position of the real-time water flow blocking point identified by the water conservancy topology network and the geological vulnerable area distribution map, and generating and executing a water conservancy facility regulation and control instruction by taking minimization of an overlapping area of the two as a target.

The invention discloses a 3D multi-body system equivariant trajectory prediction method based on multi-scale feature enhancement. The method comprises the following steps: A, obtaining trajectory data to be predicted; b, constructing a multi-scale feature enhancement model; the multi-scale feature enhancement model comprises a direction information decoupling module, a feature enhancement module, a screen type multi-scale feature updating module and a direction information coupling module; c, training the multi-scale feature enhancement model to obtain a trained multi-scale feature enhancement model; and D, performing 3D multi-body system isovariant trajectory prediction on the data to be subjected to trajectory prediction by using the trained multi-scale feature enhancement model, and finally obtaining a trajectory prediction result. According to the method, high-precision prediction of the trajectory of the 3D multi-body system can be realized, and strict constant change of the prediction process is ensured.

The invention belongs to the technical field of remote sensing image processing, particularly relates to an urban remote sensing image segmentation method and system based on bidirectional coordinate attention and multi-scale adaptive feature fusion, and provides a remote sensing image semantic segmentation neural network architecture taking an attention re-calibration module as a decoder core. Wherein the encoder path gradually extracts multi-scale feature representation through cascaded residual convolution blocks and down-sampling operation to form a feature pyramid of which the spatial resolution is reduced step by step and semantic information is enhanced step by step; the decoder path gradually recovers the spatial resolution through cascaded up-sampling and feature refining operations to generate a precise segmentation mask; a space-channel dual attention re-calibration module oriented to a decoding stage is provided, through explicit coding of space coordinate direction information and combination of global channel dependence modeling, features beneficial to semantic discrimination are adaptively enhanced in the feature fusion and resolution recovery process, and therefore segmentation precision and consistency are improved.

The invention provides an on-machine measurement path planning method for a five-axis numerical control machine tool of a large structural member, a medium, equipment and a product, and relates to the technical field of numerical control machining, the method comprises the following steps: selecting a to-be-measured surface on a three-dimensional model of the structural member, and determining a measurement point which comprises the spatial position of the measurement point and normal direction information of the to-be-measured surface at the measurement point; generating a first measurement path according to the measurement point information and the measurement path parameters, performing collision detection on the initial measurement path, judging whether interference exists between a path track and the structural member, and if interference exists, correcting a local section of the interference path to obtain a second measurement path; probe movement is simulated based on the second measurement path, probe collision detection is carried out, and if interference is found, the second measurement path of the collision area is modified to obtain a final measurement path; and generating the final measurement path into an NC program which can be identified by the five-axis numerical control machine tool, carrying out measurement work, and carrying out error compensation on actual measurement data. Manual dependence is reduced, and the measurement precision is high.

The embodiment of the invention provides an intelligent energy-saving shipborne anti-icing electric heat tracing monitoring system and method, the monitoring system comprises a plurality of electric heat tracing loops, and each electric heat tracing loop is provided with a working state acquisition module; the power distribution system supplies power to each electric heat tracing loop and gives a unique identification code to each electric heat tracing loop; the data collecting and monitoring system comprises a PLC control unit, a sensor unit and an electronic chart system, and the PLC control unit is connected with the electric heat tracing control boxes, the sensor unit and the electronic chart system and used for collecting working state information, environment temperature information, wind speed and wind direction information and ship azimuth information of the electric heat tracing control boxes; and the working state of each electric heat tracing loop is controlled. According to the embodiment of the invention, the real-time dynamic monitoring of each electric heat tracing loop of the electric heat tracing system can be realized, the control convenience and the maintenance convenience of the electric heat tracing system are improved, and the service life of the electric heat tracing system and the electric heat tracing band is prolonged.

An image processing apparatus is disclosed. The image processing apparatus includes a communicator configured to transmit or receive information to or from an intraoral scanner, a memory storing one or more instructions, and a processor configured to execute the one or more instructions stored in the memory to identify an oral image obtainment mode based on tip information about a tip included in the intraoral scanner, and obtain an oral image based on two-dimensional image data received from the intraoral scanner through the communicator and the identified oral image obtainment mode, wherein the tip information may include at least one of tip fastening direction information and tip size information.