33 results about "Deformation vector" patented technology

The invention relates to the technical field of image data processing and intelligent control, and discloses a patch board spot welding track intelligent control method and system based on visual perception. According to the method, synchronous image streams are collected through a binocular vision sensor, and epipolar correction image pairs are generated through timestamp synchronization, ROI extraction and epipolar correction processing; calculating a disparity map by adopting a stereo matching algorithm, and constructing a compensated three-dimensional point cloud model based on deformation vector field fusion multi-frame point cloud data of a radial basis kernel function; welding spot position error vectors are generated by extracting welding spot feature points and performing spatial filtering optimization; and carrying out inverse kinematics calculation on the mechanical arm by adopting a damping least square method, carrying out safety constraint optimization in combination with prospective collision risk assessment and real-time pose data, and generating a trajectory compensation instruction. According to the method, the problems of dynamic deformation compensation and motion safety in patch plate welding are solved, the submillimeter welding spot positioning precision is achieved, and the welding quality stability and the system robustness are improved.

The invention relates to a deformation monitoring method of a wind generating set blade in a running state, which comprises the following steps of: arranging a fiber grating sensor array in the blade, and collecting multi-section strain data; the strain is converted into curvature and torsion rate through a sensitivity matrix, and deflection and torsion angle are obtained through integral calculation; a laser range finder is installed in a cabin, and a space contour point cloud at the moment of blade scanning is obtained; absolute displacements of a plurality of reference sections are extracted from the point cloud, and full-field correction is carried out on the integration result of the fiber bragg grating through piecewise linear interpolation; outputting a three-dimensional deformation vector of the blade in a world coordinate system in combination with SCADA operation parameters; the deformation monitoring method of the wind generating set blade in the running state can achieve full blade coverage, accurately capture the deformation direction and deeply integrate with other running data, and has become urgent demands for industry development.

The application discloses a multi-model integrated small sample credit score offset adaptive correction system and relates to the technical field of data processing. The system comprises a divergence quantization routing decision module, which acquires target small samples and source domain samples, calculates the JS divergence of each feature dimension, and generates a decoupling mask vector to separate the original feature space into anchor features and drift features; a controlled alignment dynamic demarcation module, which performs an antagonistic feature alignment on the drift features, extracts a deformation vector, and dynamically demarcates the boundary of a decision fuzzy area based on the module length; a partial derivative projection directional correction module, which intercepts samples according to the relative position of the fluctuation boundary and the fuzzy area boundary predicted by a multi-base model, and implements nonlinear directional correction on the intercepted samples based on the dot product of the deformation vector and the partial derivative gradient; and a noise reduction verification closed-loop evaluation module, which generates a final credit score and outputs a verification index of the correction effect. The application realizes adaptive offset correction of credit scores in a small sample scenario and improves the scoring accuracy.

The present application relates to a kind of shield tunnel irregular load inversion method based on neural network agent model, belong to shield tunnel load inversion field.The method, discard the assumption of load distribution mode in traditional lining load inversion method, inversion result can reveal the irregular load distribution mode of on-site tunnel, provide technical support for the service safety of tunnel service;Inversion is based on the deformation of lining whole circumference obtained by three-dimensional laser scanning technology, significantly increase the amount of field data, with the increase of n, the dimension of deformation vector can be increased sufficiently, avoid the defect that traditional inversion method easily leads to the non-uniqueness of inversion result.

The invention discloses a multi-model integrated small sample credit score offset self-adaptive correction system, and relates to the technical field of data processing, and the system comprises a divergence quantization routing decision module which obtains a target small sample and a source domain sample, calculates the JS divergence of each feature dimension and generates a decoupling mask vector; separating the original feature space into anchoring features and drift features; the controlled alignment dynamic demarcation module performs adversarial feature alignment on the drift features, extracts a deformation vector and dynamically demarcates a decision fuzzy region boundary based on the modulus length of the deformation vector; the partial derivative projection directional correction module performs sample interception according to the relative position of the fluctuation boundary and the fuzzy region boundary predicted by the multi-base model, and performs nonlinear directional correction on the intercepted sample based on the dot product of the deformation vector and the partial derivative gradient; the noise reduction verification closed-loop evaluation module generates a final credit score and outputs a verification index of a correction effect; according to the invention, the self-adaptive offset correction of the credit score in a small sample scene is realized, and the score accuracy is improved.

The application relates to a shaft structure deformation identification method and device, equipment, a storage medium and a program product, and relates to the technical field of tunnel construction monitoring. The application can effectively detect shaft deformation. The method comprises the following steps: at multiple detection time points, positioning data generated by a UWB base station when moving and ranging data between UWB tags when moving each time are acquired; the multiple UWB tags are positioned by a least square method according to the positioning data and the ranging data, and positioning coordinates of potential deformation positions at the multiple detection time points are obtained; for any potential deformation position, a deformation vector is constructed based on the positioning coordinates of any adjacent detection time points, and a time sequence of the deformation vectors of the multiple potential deformation positions is obtained; deformation risk judgment is performed on the multiple potential deformation positions based on the time sequence of the deformation vectors, and when there is deformation risk at any potential deformation position, deformation feature recognition is performed based on a two-dimensional image and three-dimensional point cloud data of a target deformation position with the deformation risk.

A medical use image processing method executed by a medical use image processing device including a processor, the method includes a reception step of receiving input of a first medical use image actually captured in a first posture, an image generation step of generating a second medical use image, in which the same portion as the first medical use image is imaged in a second posture different from the first posture, from the first medical use image in a pseudo manner, in which the second medical use image is generated by using a deformation vector field that converts the first medical use image into the second medical use image, and a modality conversion step of converting the second medical use image generated in the pseudo manner into a third medical use image having a modality different from the second medical use image in the pseudo manner.

The invention discloses a multi-process thermal mechanical deformation monitoring and compensating system for an engine support, and relates to the technical field of image monitoring. The system firstly collects visible light and infrared data to generate a fusion image, and recognizes the chip shielding state of the surface of a workpiece; if yes, driving a physical cleaning device to perform directional cleaning until a judgment logic is met, and generating a hot historical log containing cleaning process data; calculating a current three-dimensional full-field deformation vector of the workpiece in combination with the recovered pure surface temperature field and a thermal history log; and finally extracting the displacement and rotation amount of the key feature points, and generating a machine tool coordinate system space transformation instruction for compensation. According to the method, the problems of infrared temperature measurement distortion caused by inner cavity chip accumulation under the minimal quantity lubrication working condition and thermal lag prediction deviation generated by internal heat accumulation after chip removal are solved.

This invention discloses a method and system for analyzing and outputting TCM acupuncture cases based on machine learning, belonging to the field of medical image processing technology. The method acquires facial image sequences from two pathological cycles of the subject, extracts spatiotemporal feature maps, and obtains local receptive field feature matrix pairs through sliding windowing. It calculates the cross-cycle joint entropy gradient and its matrix, and generates a two-dimensional deformation vector using optical flow. Using the coordinates of the first cycle as a reference, the deformation vector is set to zero when the joint entropy gradient does not exceed the rigidity threshold; otherwise, the joint entropy gradient matrix is ​​used as a weight matrix through a sigmoid activation function, and a Hadamard product is performed with the deformation vector to obtain a denoised deformation vector. The denoised vectors are superimposed to obtain corrected pixel coordinates, which are mapped to the target localization coordinate matrix, and the acupuncture case analysis results are output. This scheme achieves precise spatial decoupling of local tissue evolution, eliminates the contamination of rigid regions by global deformation, and significantly improves the robustness and accuracy of cross-cycle acupuncture target localization.

The present application provides a kind of respiratory motion compensation method based on manifold clustering and weighted sparse, comprising the following steps: collecting the 4D-CT image of population data set and the 3D-CT image of individual subject expiration-inhalation two phases;Obtain the deformation vector field motion characteristics of population data set and individual subject by image registration;Obtain the sub-population data set similar to individual subject motion characteristics based on manifold clustering;The quantitative relationship between individual and sub-population motion characteristics is established by using weighted sparse method, and individual 4D-CT image is constructed;Obtain the internal deformation vector field and external skin surface motion of individual subject, and construct individualized respiratory motion compensation model to realize real-time prediction of lung image.The present application can reduce the radiation damage of the subject while ensuring accurate prediction of lung movement, and is expected to break through the limitations of existing respiratory motion compensation methods, can real-time image prediction for whole lung, reduce the influence of respiratory motion on lung tissue.

This invention relates to the field of intelligent compensation technology, specifically to an online compensation method for wear of belt pulley machining tools that integrates AI visual detection. The method includes the following steps: acquiring tool wear images and cutting force data, inputting them into a convolutional neural network to analyze the wear degree; detecting node stress ratios to group and form a stress heterogeneous mapping; monitoring displacement and stress angles to identify and optimize the trajectory of inverted nodes; decoupling deformation vectors to calculate correlation coefficients and classify to obtain a set of deformation compensation parameters for belt pulley machining tools; and calculating compensation prediction weighted summaries to generate online compensation commands for belt pulley machining tools. In this invention, visual recognition and multi-dimensional stress analysis are integrated to achieve tool wear assessment; stress distribution mapping presents stress-displacement correlation; dynamic trajectory optimization balances force; deformation is adaptively adjusted; nonlinear deformation and overall trend are synchronously corrected; compensation parameters are self-generated; deformation trajectory and machining path are coordinated; errors are dynamically suppressed; dimensional offsets and surface fluctuations are mitigated; and machining stability and response time are improved.

The application relates to the technical field of dynamic modeling, and discloses a dynamic modeling method and system for a 3D home scene, which comprises the following steps: acquiring real-time interactive displacement vector data; when the capture frequency of the real-time interactive displacement vector data exceeds a fixed refresh rate threshold, fusing into a weighted mixed deformation vector through a cyclic temporary storage queue; calculating a local vertex attribute matrix correction value according to the weighted mixed deformation vector, and performing matrix internal superposition transformation on the local vertex attribute matrix correction value and a time sequence topological offset matrix to complete time sequence difference compensation correction; and transmitting the local vertex attribute matrix correction value into a sub-buffer area overwrite interface through a sparse vertex array slice stored in an interleaved structure to perform local in-situ overwrite writing, and updating a grid topological form, wherein the application limits the recalculation cost within a local slice range, reduces the throughput pressure of a display memory bus, eliminates time sequence lag, and avoids picture flickering.

A method and system for reconstructing motion compensated nuclear images of an object and an arrangement for acquiring nuclear images and a computer program for performing the method are provided. The reconstruction method comprises receiving nuclear image data for a plurality of motion states, reconstructing the data into images for each motion state and computing a deformation vector field for each state for mapping the images onto a reference motion state. Computing the deformation vector field comprises providing an initial vector field, defining at least one rigid region of the object, incorporating the rigid region into the initial deformation vector field and computing the deformation vector field with the incorporated rigid region. The method further comprises mapping the reconstructed images of each motion state onto the reference state using the deformation vector field and combining the mapped nuclear images into a motion compensated nuclear image.

A medical use image processing method according to an aspect of the present invention is a medical use image processing method executed by a medical use image processing device including a processor, the medical use image processing method including causing the processor to execute a reception step of receiving input of a first medical use image actually captured in a first posture, and an image generation step of generating a second medical use image, in which the same portion as the first medical use image is imaged in a second posture different from the first posture, from the first medical use image in a pseudo manner, in which the second medical use image is generated by using a deformation vector field that converts the first medical use image into the second medical use image.

The invention relates to the technical field of image segmentation, in particular to a lung tumor CT image automatic segmentation method based on deep learning. The method comprises the following steps: firstly, acquiring three-dimensional CT images of different breathing time phases and a tumor area in the first image; then acquiring a rigid transformation model of lung movement; marking points are determined in a preset range of the tumor area, marking points in each non-first image and deformation vectors of the marking points are determined, stable marking points are screened out, deformation vectors of the unstable marking points are corrected, and then each three-dimensional CT image and correction positions of the marking points in the three-dimensional CT image are input into a trained segmentation model to obtain a segmentation model; and obtaining a lung tumor segmentation result. According to the method, the respiratory movement of the lung tissue is decomposed and analyzed, the movement changes of the mark points in the tumor area in different respiratory time phases are accurately evaluated, then automatic segmentation is carried out based on the segmentation model of deep learning, and the lung tumor CT image segmentation effect based on deep learning is improved.

This disclosure relates to the field of computer graphics technology, and provides a method, apparatus, device, medium, and program product for clothing adaptation of virtual characters. The method for clothing adaptation of virtual characters includes: first, establishing a three-dimensional spatial mesh containing a source virtual character model and a target virtual character model with the same topological structure, as well as a source clothing model; then, based on the positional correspondence between the source and target virtual character models, and combined with smoothing constraints to limit the difference in deformation vectors between adjacent mesh points, calculating the deformation vector of each mesh point in the three-dimensional spatial mesh; finally, using the deformation vector of each mesh point, mapping and transforming the source position of each vertex in the source clothing model to the target position, thereby generating a target clothing model adapted to the target virtual character model. This disclosure can effectively maintain the geometric characteristics of the clothing model, reduce distortion during the adaptation process, and improve visual effects.

The invention discloses a power transmission line tower deformation intelligent monitoring method and system based on visual identification, and relates to the technical field of power transmission line state monitoring, and the method comprises the steps: collecting historical normal observation frames, constructing a historical main track and a flow frame, defining a deformation vector, and constructing a historical mode matrix; establishing an optimized objective function based on a historical pattern matrix and a deformation representation vector, solving by using ADMM to obtain a sparse coefficient vector, calculating a reconstruction error and converting the reconstruction error into error energy, calculating an error normalization shape, obtaining an InSAR time sequence deformation sequence and calculating a normalization risk score, and outputting a deformation determination result through DBSCAN clustering; structural damage sensitive features are generated through a curvature residual error and multi-mother wavelet combined spectrum matrix, the multi-scale recognition precision of tiny local damage and the overall deformation trend is improved, steady-state characterization of a tower image in a complex environment is achieved by constructing a sequence reference frame, and the robustness of deformation detection is improved.

This application relates to the field of port foundation deformation monitoring technology, and discloses a risk prediction method for soft soil foundation scenarios. The method includes: acquiring deformation data and usage data at several sampling times within several grid areas; calculating the data difference between adjacent deformation data in each grid area; combining the data difference from a specific historical time to the current time and the grid position coordinates within each grid area into a deformation vector; clustering multiple deformation vectors to obtain one or more clusters; obtaining the correlation between historical overall usage data and historical overall deformation data; obtaining future overall usage data; determining the future overall deformation data corresponding to the future overall usage data based on the correlation; and determining the risk level corresponding to the future overall deformation data. This method can be used to predict the deformation risk of soft soil foundations in use. This application also discloses a risk prediction device and system for soft soil foundation scenarios.

The invention provides a non-contact ground fracture monitoring method, device and system fusing multi-source data, and a storage medium. The method comprises the steps of obtaining a two-dimensional visible light image sequence and a three-dimensional depth image sequence of a to-be-monitored ground fracture region; obtaining a two-dimensional plane time sequence deformation vector and a three-dimensional deformation feature sequence of the edge key points of the ground fracture; obtaining a ranging data sequence of ground fracture edge key points; dynamically determining weights of the two-dimensional plane time sequence deformation vector, the three-dimensional deformation feature sequence and the ranging data sequence; establishing a transfer function between the fusion data matrix and a ground fracture target deformation result; and forming an observation data conditional probability based on the sensor measurement error and the corresponding weights of the two-dimensional plane time sequence deformation vector, the three-dimensional deformation feature sequence and the ranging data sequence, and performing maximum likelihood estimation on the observation data conditional probability to obtain a ground fracture target deformation result. By adopting the monitoring method provided by the invention, the ground fracture deformation can be monitored more accurately.

By analyzing crustal deformation data, we can clearly identify precursory anomalies such as pre-slips that precede major earthquakes, and use this data to predict earthquakes and estimate the optimal epicenter location from among potential epicenters. [Solution] The crustal deformation vector calculation unit calculates the difference vector of the change in position information at each observation station at a fixed interval such as 5 minutes, and calculates this as the crustal deformation vector for each observation station. The correlation value calculation unit performs processing based on the crustal deformation vectors for each observation station output by the crustal deformation vector calculation unit. The correlation value calculation unit selects one station as the central observation station and selects several observation stations in its vicinity as surrounding observation stations. In a certain situation, the correlation value calculation unit can select any number of surrounding observation stations from among observation stations whose distance from the central observation station is within a predetermined distance.

The present application relates to a target region position recognition method and system based on image data, a planning image with a target region contour and a current image to be registered are acquired, preliminary alignment is performed; an initial deformation vector field is calculated by taking the planning image after preliminary alignment as a floating image, a local volume change graph and a tissue elastic modulus graph are generated; a spatial variation regularization term is determined, a deformation penalty is applied to a voxel with a local volume change rate lower than a first threshold value and an elastic modulus higher than a second threshold value by using the spatial variation regularization term, and a final deformation vector field is obtained based on the deformation penalty; an initial current target region contour is taken as a starting point, contour evolution is performed in a gradient field of the current image to obtain an optimized contour, and a volume change rate between the optimized contour and the initial current target region contour is calculated; when the volume change rate exceeds a third threshold value, the smoothing strength of the optimized contour is increased to obtain a target region position.

The present application relates to the technical field of intelligent medical monitoring, in particular to a burn wound healing process real-time monitoring method and system, comprising the following steps: collecting oxygen saturation data to construct a microcirculation blood flow feature set, performing texture tracking on a wound image to construct a global deformation vector field, fitting an edge line and combining the deformation field to obtain a net proliferation migration vector field, constructing a recursive feature matrix and calculating a laminar flow property and a capture time index, and outputting an evaluation result, in the present application, by combining microcirculation blood flow features and deformation field mapping, background motion interference is eliminated, net proliferation migration behavior is accurately quantified, a recursive feature matrix is constructed by mapping time series signals to phase space, laminar flow property and capture time index in state evolution are identified, multi-dimensional state determination is performed by fusing physiological blood flow parameters and morphological contraction coefficients, non-contact objective quantitative analysis of the healing process is realized, multi-dimensional pathological grading standards are established, and the dynamic trend of tissue remodeling is accurately reflected.

This application relates to the field of intelligent manufacturing technology, and in particular to a method, system, and storage medium for controlling the deformation of aluminum frame processing based on data monitoring. The method includes: acquiring surface curvature and deformation data of the aluminum frame through sensors, generating a curvature distribution map and deformation vector field after image processing; analyzing the data based on a neural network model to identify stress distribution and potential defects, and determining speed adjustment requirements in conjunction with equipment delay information; if the requirements exceed a threshold, generating a dynamic speed curve and extracting key parameters to update equipment control; the system monitors the execution effect through feedback loops, and if the indicators do not meet the standards, it backtracks the data, optimizes model parameters, drives the equipment to complete processing through improved instructions, and collects subsequent data to evaluate the overall quality; this application improves the processing quality of aluminum frames.

The invention relates to the technical field of building monitoring, in particular to a tower-type historic building deformation detection and analysis method and device, and the method comprises the following steps: S1, scanning a target tower-type historic building, and obtaining the three-dimensional point cloud data of each period; s2, establishing a reference coordinate system by taking the bottom end center point of the reference central axis as an original point and taking the central axis direction as a Z axis; s3, performing overall registration on any monitoring period point cloud and the reference period point cloud; s4, extracting an outer contour line of each horizontal section intersecting with the point cloud in the monitoring period; s5, calculating a radial offset, and generating a deformation vector set of the section; and S6, synthesizing the deformation vector sets of all the elevation profiles, and generating a three-dimensional deformation field graph of the whole ancient tower. According to the method, the multi-period point cloud is registered through the unified coordinate system, the contour deformation of each section is extracted, the three-dimensional deformation field graph is constructed, and accurate detection and visual analysis of the structural deformation of the tower-type historic building are realized.

The invention relates to the technical field of information, and provides a cargo stacking management method and system, electronic equipment and a storage medium. The method comprises the following steps: monitoring the deformation of a cargo to determine the geometric shape change of the cargo; according to the geometric shape change, updating a stacking geometric model used for representing the spatial relationship between the cargoes; and based on the updated stacking geometric model, a stacking scheme of the goods is adjusted. According to the invention, by monitoring the deformation track of the goods caused by environmental factors in real time and dynamically updating the spatial relation matrix and the stacking geometric model, the crossing from static planning to dynamic adaptive management is realized; a deformation vector field is utilized to correct coordinates of a contact surface in real time, and a particle swarm optimization algorithm is combined to optimize an embedding depth in a mechanical level, so that remapping of a layout grid can be automatically triggered when a cargo contracts or expands in a physical dimension, and the accuracy of the physical dimension of the cargo is improved. And the long-term geometric stability and safety of goods stacking in a complex environment are improved.

The invention relates to the technical field of port foundation deformation monitoring, and discloses a risk prediction method for a soft soil foundation scene. The method comprises the following steps: acquiring deformation data and use data at a plurality of sampling moments in a plurality of grid regions; calculating a data difference value of adjacent deformation data in each grid region; combining the data difference value from the specific historical moment to the current moment in each grid region and the grid position coordinates into a deformation vector; clustering the plurality of deformation vectors to obtain one or more clusters; obtaining an association relationship between the historical overall use data and the historical overall deformation data; obtaining future overall use data; determining future overall deformation data corresponding to the future overall use data according to the association relationship; and determining a risk level corresponding to the future overall deformation data. By adopting the method, deformation risk prediction of the soft soil foundation in a use state can be realized. The invention further discloses a risk prediction device and system for the soft soil foundation scene.

The invention provides a hyperspectral remote sensing image splicing method based on deformation transmission along a splicing line, and the method comprises the steps: S1, constructing a function in an overlapping region of adjacent images based on the spectral similarity information of the images for a registered hyperspectral image, and solving through a graph cut model to obtain an optimal splicing line; s2, in two adjacent hyperspectral images, performing one-dimensional feature point extraction and matching along a splicing line based on spectral gradient information; and S3, based on the matched one-dimensional feature points on the splicing line of the two adjacent hyperspectral images, obtaining a deformation vector describing geometric position difference between the feature points, solving a deformation vector of any pixel point in the overlapping region of the adjacent images through a geometric relationship, and realizing smooth deformation transmission from the feature points on the splicing line to the global image. And extracting and matching the one-dimensional feature points in the two registered images along the splicing line to obtain the geometric deformation relationship of the typical surface feature boundary on the splicing line.

The invention relates to the technical field of artistic design interaction, and discloses an artistic design interaction experience system, which comprises a gesture sensing module for acquiring hand motion data and muscle activity electric signals in real time, extracting motion features and identifying non-standard gestures deviating from standards; the information acquisition module is used for collecting creation context information such as a virtual object state and operation history when a non-standard gesture is recognized; the intention inference and instruction correction module is used for comparing and correcting non-standard gestures in combination with contexts to obtain an operation instruction conforming to the creation intention; the data recording and associating module is used for recording non-standard gesture original data and deformation vector data generated by a correction instruction, and associating to form an operation traceability chain; the backtracking correction module provides a visual interface and a deformation backtracking tool, and corrects a local deformation area of the virtual model based on a traceability chain; according to the method and the device, the problems of inaccurate non-standard gesture recognition, force feedback interruption and difficulty in accurate backtracking and fine model deformation correction in the prior art are effectively solved.

This invention relates to the field of facial expression recognition technology, specifically to an intelligent assessment method for anesthesia and analgesia status based on facial expression recognition. The method includes: for the user to be assessed, obtaining a mask overall translation reference vector and a facial pixel movement vector matrix; stripping the global translation component from the facial pixel movement vector matrix and generating a local contraction deformation matrix; based on the mask overall translation reference vector, performing linkage component removal processing on the local deformation vectors of each grid node in the local contraction deformation matrix to separate independent contraction components and constructing a facial independent contraction feature matrix; inputting the facial independent contraction feature matrix into a convolutional neural network to extract the classification probabilities of each preset category of facial movements, and performing weighted fusion processing to determine the analgesia status assessment index of the user to be assessed, which indicates that the user only responds to independent facial muscle contractions induced by the patient's central nervous system, effectively improving the accuracy of the anesthesia and analgesia status assessment results.

Provided are an image processing method, an image processing device, a program, and a trained model that can suppress calculation resources required for registration between a plurality of images.An image processing method that is executed by one or more processors includes: allowing the one or more processors to acquire feature maps of each of a plurality of images; and allowing the one or more processors to calculate a deformation vector field from a combination of the feature maps of each image. The one or more processors generate the feature maps respectively from the plurality of images using a first neural network and input the generated combination of the feature maps to a second neural network to calculate the deformation vector field.