21results about How to "Precise registration" patented technology

The invention discloses a tunnel crack detection method and a tunnel crack detection system, and belongs to the field of tunnel safety detection.The tunnel crack detection method comprises the steps that a tunnel inspection robot is utilized to conduct inspection in a tunnel to obtain complete three-dimensional data in the tunnel so as to construct a three-dimensional map in the tunnel, and the position of each crack area in the three-dimensional map is determined; determining an inspection tour area corresponding to each crack area and executing path planning, and generating an optimal inspection tour path of the unmanned aerial vehicle; the optimal inspection path enables the unmanned aerial vehicle to sequentially pass through each inspection area and then leave the tunnel after entering the tunnel under the condition that obstacle avoidance conditions are met, and the total time is shortest; enabling the unmanned aerial vehicle to move according to the optimal inspection path, and scanning the crack region by using the depth camera when the unmanned aerial vehicle moves to one inspection region so as to obtain fine three-dimensional data of the crack region; and performing analysis processing on the fine three-dimensional data of the crack area to complete tunnel crack detection. According to the invention, the efficiency and precision of long tunnel crack detection can be improved.

The invention discloses a tooth three-dimensional digital model registration and tooth neck line extraction method based on oral scanning and CBCT, and relates to the field of tooth three-dimensional digital model registration. Comprising the following steps: respectively acquiring three-dimensional oral images of oral scanning and CBCT (cone beam computed tomography), and automatically segmenting three-dimensional structures of dental crowns and teeth by utilizing pre-trained V-Net and ToothNet image segmentation models. Alignment and registration of two modal images are realized through a feature extraction network, an information exchange network and a rigid transformation network. The feature extraction network extracts local and global features by using far-point sampling and a dynamic graph convolutional neural network, the self-attention and cross-attention module performs feature enhancement and information exchange between the same mode and different modes, and based on a registration result, a tooth neck line is further extracted and optimized. According to the method, the tooth three-dimensional digital model can be quickly and accurately registered, a high-precision tooth neck line extraction result is provided, the manpower and time cost of a medical institution can be reduced, and the medical efficiency is improved.

The invention provides a two-stage registration method and device suitable for a full-length image of a lower limb, a terminal and a medium. The method comprises the steps of obtaining a to-be-registered CT image and a label of the lower limb, and a target X-ray image and a label of the lower limb; inputting the to-be-registered CT image and the label, and the target X-ray film image and the label into an affine transformation network to generate an affine transformation matrix for performing rough registration on the to-be-registered CT image; and inputting the CT image and the label output in the coarse registration stage, and the target X-ray film image and the label into a deformation field registration network to generate a deformation field for fine registration based on the coarse registration result of the CT image to be registered. According to the method, the deep learning technology is utilized, a brand-new 2D / 3D medical image registration algorithm is constructed, and full-automatic cross-modal registration of lower limb full-length horizontal position CT and standing position X-ray films is achieved for the first time. The research result of the invention can provide effective assistance for preoperative accurate planning, assist the design work of an HTO operation guide plate, and improve the clinical curative effect of early and middle knee OA.

The invention discloses a neural field guided point cloud registration method and device, electronic equipment and a storage medium, and belongs to the technical field of point cloud data processing, and the method comprises the following steps: building a neural network for guiding point cloud registration by using a neural field, inputting a source point cloud, and outputting rotation angles and displacements of x, y and z axes; expressing the target point cloud in the form of a neural field by using a small neural network; matrix orthogonalization constraint setting is carried out on the registered neural network; introducing the constructed target point cloud neural field to construct a loss function for the registration neural network; training a registration neural network and predicting a rotation angle and displacement by taking source point cloud data as input, obtaining a rotation matrix by utilizing matrix orthogonalization constraint, and continuously approaching an optimal transformation matrix through back propagation under the guidance of a loss function; and performing registration of the source point cloud and the target point cloud by using the optimal transformation matrix. According to the method, the neural field is introduced into point cloud registration, so that rapid and accurate point cloud registration is realized.

The invention provides a vehicle positioning method based on laser radar point cloud data. The method comprises the following steps: acquiring a source point cloud and a target point cloud; using an improved point-to-surface iterative nearest point algorithm to decentralize the source point cloud and the target point cloud, calculating a rotation matrix and a translation vector according to decentralized coordinates of the source point cloud and the target point cloud, and combining input disturbance and Lie algebra to obtain an initial registration conversion formula of the source point cloud; performing Gaussian Newton iteration on the initial registration conversion formula to obtain a registration conversion formula of the source point cloud, and performing registration on the source point cloud and the target point cloud to obtain registration point cloud data; and constructing a map by using the registered point cloud data, and positioning the vehicle. According to the invention, the technical problem that the algorithm precision is not high when vehicle positioning is carried out through point cloud data in the prior art can be solved.

The invention provides a security situation awareness method and system based on a three-dimensional GIS and a panoramic video, and the method comprises the steps: constructing a three-dimensional scene model of a monitoring region on a three-dimensional GIS platform, accessing a plurality of cameras and an environment sensor, and achieving the space-time alignment and coordinate unification of a video stream and multi-source sensing data; projecting a video frame to a three-dimensional scene based on internal and external parameters of a camera to generate a panoramic video picture, and performing detection and cross-camera tracking on pedestrians and vehicles by adopting a target detection model to obtain a target three-dimensional track; predicting a target position and local density change by using a time sequence prediction model, judging anomalies including congestion, retention and retrograde driving, and sending out an early warning; camera deployment and cooperative tracking are planned through a point location optimization algorithm, and monitoring blind areas are reduced; the system comprises a data acquisition module, a space-time alignment module, a three-dimensional mapping module, a target detection and tracking module, a situation analysis module, a point location planning module and a situation display module, and can improve the automation and precision of security situation awareness in a complex scene.

The present application provides a registration device, system and method for a roll-to-roll printer. The registration device is used to register the graphic information printed multiple times on the printing medium by the roll-to-roll printer, and includes: a support platform, which is arranged facing the paper feed of the roll-to-roll printer; a bracket, which is arranged on the support platform along the width direction of the paper feed; an image acquisition device, which is arranged on the bracket, and the image acquisition device can move on the bracket along the width direction to capture the image of the printing medium, so as to determine whether the graphic information printed each time can be registered. The technical solution of the present application can replace the operator's manual judgment method to determine whether the graphic information printed each time can be registered under the limitations of the structural characteristics of the roll-to-roll printer and its characteristics when it is continuously running and processing printing media, so as to ensure that the graphic information printed each time can be registered with each other.

The invention discloses a binocular vision system image registration calibration method. The method comprises the following steps: 1) constructing a binocular image acquisition system; 2) designing a special target for binocular system calibration; 3) calibrating the R and T relation of the binocular camera by using a special target to realize coarse registration; 4) actually acquiring a binocular image of the to-be-measured piece, and extracting edge contour information of the binocular image; and 5) realizing high-precision registration of the binocular image from coarse to fine according to an optimization strategy. According to the binocular vision system image registration calibration method, an advanced image processing algorithm, a precise optical measurement principle and an intelligent optimization strategy are ingeniously integrated, and remarkable acceleration of the calibration process and remarkable improvement of the calibration precision are achieved.

The invention relates to the technical field of medical imaging, and discloses a lower limb vein 3D imaging method, which comprises the following steps: identifying an imaging range of a lower limb vein patient, and collecting a lower limb slice image of the lower limb vein patient; analyzing the pixel distribution condition of the tissue type, and identifying a venous blood pixel intensity factor of the patient with the lower limb vein; constructing a pixel integral and an image of the corrected image, performing image segmentation processing on the corrected image to obtain a segmented image, and determining a vein blood vessel of the patient with the lower limb vein; analyzing the influence degree of the feature type on the lower limb vein patient, and performing dimension reduction processing on the vein features to obtain key vein features; and according to the key vein features, the pixel integral and the image, constructing a lower limb vein 3D image of the lower limb vein patient, analyzing an imaging effect of the lower limb vein 3D image on a vein blood vessel, and based on the imaging effect, outputting a 3D imaging result of the lower limb vein patient. The three-dimensional structure and the spatial relation of the lower limb veins can be visually displayed, and the accuracy of diagnosis results is improved.

The present application relates to an image registration method, device, computer equipment and storage medium. The method comprises: performing different image transformation processing on the image to be registered according to multiple transformation methods to obtain multiple images to be matched; obtaining a reference image and determining at least two local feature areas in the reference image; determining candidate areas that match each local feature area respectively from each of the images to be matched, and filtering out a target matching image from the images to be matched according to the candidate areas; determining the target transformation parameters between the reference image and the image to be registered according to the candidate areas that match the at least two local feature areas in the target matching image; registering the images to be registered based on the target transformation parameters to obtain the registered images. The use of this method can improve the accuracy of image registration.

The application discloses a kind of based on deep feature fusion's cable membrane structure digital twin model reconstruction method, including survey entity cable membrane space structure, arrangement acquisition station, acquisition cable membrane space structure three-dimensional point cloud data;The three-dimensional point cloud of cable membrane space structure is preprocessed and constructs dataset;RPMNet network architecture is improved, and spatial attention mechanism method is applied to the point cloud feature extraction layer of RPMNet network architecture;The three-dimensional point cloud of cable membrane space structure is finely registered;Based on the three-dimensional registration point cloud of cable membrane space structure, fine model reconstruction is carried out;Deviation analysis is carried out to the fine model reconstruction of cable membrane space structure and theoretical design model.The method realizes the accurate registration of three-dimensional point cloud data of cable membrane space structure, completes the high-fidelity digital twin model reconstruction of entity cable membrane structure space form, has the characteristics of high efficiency, low cost and wide coverage, has good applicability to cable membrane space structure, and can be widely applied to engineering practice.

The cell image registration method provided by the embodiment of the present invention includes: inverting the brightness of the background and target of the first bright field cell image to obtain a second bright field cell image, and using the first fluorescent cell image as the second fluorescent cell image; corroding the one with more pixels covered by the target in the second bright field cell image and the second fluorescent cell image to obtain a third bright field cell image and a third fluorescent cell image; and registering the first bright field cell image and the first fluorescent cell image according to the phase correlation method, the third bright field cell image and the third fluorescent cell image. The embodiment of the present invention first makes the directions of the relative brightness of the background and the target in the bright field cell image and the fluorescent cell image the same, and then corrodes one of the bright field cell image and the fluorescent cell image, and then uses the phase correlation method for registration after the two-step preprocessing, so that the corresponding target centroids in the bright field cell image and the fluorescent cell image are aligned, and the registration is more accurate and stable.

The invention relates to a registration method for determining a position and orientation of an object relative to a position detection system, wherein the method comprises the following steps: - capturing an image of a surface of the object with an image sensor unit comprising at least one motion sensor and / or at least one position sensor, - determining a capturing position of the image sensor unit in a coordinate system of the position detection system by processing motion sensor signals provided by the motion sensor and / or position signals provided by the position sensor, - generating a surface model of the object photogrammetrically from the captured image, and - transforming the surface model generated photogrammetrically into the coordinate system of the position detection system.

The present invention provides a two-dimensional and three-dimensional image registration method, system, device, medium and program product, which relate to the field of intelligent medical care. The method includes: obtaining intraoperative anteroposterior and lateral two-dimensional images of the subject with a registration ball; selecting feature points in the two-dimensional image and obtaining the two-dimensional coordinates of the feature points in the two-dimensional image; obtaining the three-dimensional coordinates of the feature points in the world coordinate system based on the two-dimensional coordinates; obtaining the three-dimensional coordinates of the feature points in the preoperative three-dimensional model; pairing the feature points in the three-dimensional model with the feature points in the world coordinate system to obtain the rough position relationship of the three-dimensional model under the C-arm machine; based on the rough position relationship, automatically extracting the feature points in the two-dimensional image and the preoperative three-dimensional model for iterative optimization, calculating the precise position of the preoperative three-dimensional model under the C-arm machine, and completing the registration. The present application simplifies the operation and reduces the computing resources by first roughly registering to give an initial posture with a small error, and then performing precise registration based on the initial posture as the initial iteration value.

This invention discloses a block-based dynamic registration and quantization method and apparatus for vascular video. The invention includes: firstly, acquiring vascular signals from a target region; generating a dynamic vascular video based on the vascular signals; then, obtaining the positions of motion artifacts in the vascular images of each corresponding frame in the dynamic vascular video; dividing each vascular image with motion artifacts in the dynamic vascular video into several vascular image blocks; and also treating each vascular image without motion artifacts as a single vascular image block; next, acquiring a reference image; registering all image blocks to the reference image; and then quantifying the dynamic changes of the blood vessels, ultimately obtaining the dynamic change curve of the blood vessels over time. This invention achieves accurate registration of vascular dynamics videos containing motion artifacts while preserving the dynamic change information of the blood vessels, thus realizing precise quantification of the vascular dynamic response.

The invention provides a method for tracking the trajectory of a moving object in a transformer substation based on a three-dimensional digital model, and relates to the technical field of safety management and control, the method is applied to a server of a transformer substation management and control system, the transformer substation management and control system comprises the server, a plurality of controllers and a plurality of laser point cloud sensors, and the server is connected with the plurality of controllers simultaneously. Each controller is connected with a laser point cloud sensor. The method comprises the following steps: acquiring first position data transmitted by a first controller; according to a preset transformer substation space structure, calculating the probability that the moving target moves to the scanning area of each sensor; if it is determined that the probability that the moving target moves to the second scanning area is larger than or equal to a preset threshold value, an acquisition instruction is sent to a second controller; receiving change data transmitted by the second controller; and generating a continuous motion state of the three-dimensional moving target according to the change data. According to the invention, the data volume transmitted during trajectory tracking of the moving object in the transformer substation based on the three-dimensional digital model can be effectively reduced.

The invention provides a cable three-dimensional imaging method based on tensor coupling and physical constraint, and the method comprises the steps: carrying out the tensor decomposition of terahertz time-domain spectral data and X-ray projection data of a cable, and obtaining a core tensor and a modal basis matrix; registering terahertz three-dimensional imaging data and X-ray three-dimensional imaging data of the cable based on the core tensor and the modal basis matrix; according to the registered terahertz three-dimensional imaging data and the registered X-ray three-dimensional imaging data, generating three-dimensional potential distribution of the cable by using a pre-constructed joint optimization objective function for physical constraint; and based on the three-dimensional potential distribution, generating a dynamic weight through a double-branch convolutional network, and according to the dynamic weight, performing restoration by adopting a weighted difference value and a context sensing network to obtain a three-dimensional image of the cable. Therefore, deep collaborative information among different modal data is fully mined, and the structural integrity and detail definition of the image are remarkably enhanced.

The invention relates to a domain name registration method, and belongs to the technical field of domain name registration, and the method comprises the steps: 1, carrying out the grouping processing of domain names to be registered according to a grouping formula, and calculating the time range of the deletion and release of the domain names; 2, registering in a domain name deletion release period; 3, after registration is finished, the data of the deleted and released domain names in the current period are reset; and step 4, optimizing the grouping formula according to a redisk result. Through the grouping formula, the to-be-registered domain name can be accurately registered after the domain name is released, the error is small, and the efficiency and the registration success rate are higher.

The invention discloses a registration method and device for fiber bundle tracking data and a structure image and a medical system. The method comprises the steps that an anatomical structure image and an initial fiber bundle tracking image obtained by a target part of a patient through nuclear magnetic resonance imaging (MRI) equipment are obtained; based on the image data of the anatomical structure image, performing transformation processing on the image data of the initial fiber bundle tracking image to obtain a first affine transformation matrix and a transformed initial fiber bundle tracking image; the first affine transformation matrix is used for enabling the initial fiber bundle tracking image and the anatomical structure image to partially coincide in the same coordinate system; packaging the image data of the transformed initial fiber bundle tracking image to obtain a packaged fiber bundle tracking image; and registering the packaged fiber bundle tracking image and the anatomical structure image according to a preset registration rule to obtain a registered image.

The invention relates to a power equipment image registration method based on deep learning, and the method comprises an image feature extraction step which comprises a visible light image feature extraction branch; thermal imaging image feature extraction branches; extracting branches of visible light image features with marked edges; a step of applying a cross attention mechanism: obtaining first interaction information transmitted from the feature map of the thermal imaging image to the feature map of the visible light image and second interaction information transmitted from the feature map of the visible light image to the feature map of the thermal imaging image; in the feature learning step, the image feature map and the interaction information are input into an image matching learning network, and a matrix conversion result is obtained. According to the method, correlation coefficients of different images are learned, a learning network is formed for feature recognition, and finally accurate registration of the images of the power equipment is achieved. According to the method, the image registration of the power equipment is realized, and the method has the advantages of high registration precision, wide applicability and the like, and can be widely applied to the field of safety monitoring of the power equipment.

The invention relates to a loopback detection method and device and a vehicle, and the method comprises the steps: generating a multi-resolution space division tree based on stored point cloud data, the multi-resolution space division tree comprises at least one tree node, the tree node stores point cloud data, and the multi-resolution space division tree is a multilayer quadtree or octree; a point cloud frame is received, a coordinate set of to-be-registered point cloud points in the frame is acquired, a target underlying tree node is acquired based on the coordinate set, and the target underlying tree node comprises at least one coordinate in the coordinate set; judging whether the overlapping degree of the frame and the target underlying tree node meets a threshold value or not; and if the frame meets the threshold value, registering the frame and the multi-resolution space division tree, sequentially taking the optimal rotation translation parameter obtained by the tree nodes of the upper layer as the initial value of the tree nodes of the lower layer from the top layer to the bottom layer, and calculating the optimal rotation translation parameter of the tree nodes of the lower layer until the tree nodes of the bottom layer or registration fails.