45 results about "Volume reconstruction" patented technology

The invention provides a magnetic resonance image self-supervision super-resolution reconstruction method and system based on a prior guide diffusion model. The method comprises the steps of image preprocessing, pseudo-pairing data set construction, prior condition construction, model training, model testing and three-dimensional isotropic volume reconstruction. According to the method, the dependence on a real high-resolution image is remarkably reduced by constructing a self-supervised pseudo-pairing mechanism, and high-quality reconstruction from anisotropic MRI to an isotropic image is realized; meanwhile, by introducing multiple medical priori including high-frequency residual, edge structure and region-of-interest guidance, detail retention and structural boundary definition of the image are effectively improved, the convergence speed and robustness of the model are improved while anatomical fidelity is guaranteed, and the method is suitable for large-scale popularization and application. And a technically feasible and clinically applicable solution is provided for high-resolution reconstruction of medical images.

The invention belongs to the technical field of medical image reconstruction, and particularly relates to a CT reconstruction system and method based on discrete Gaussian representation. According to the technical scheme, the method comprises the steps that a Gaussian function set is constructed, and a three-dimensional voxel space is directly represented through learnable Gaussian parameters; carrying out parallel aggregation on Gaussian function contribution values to a three-dimensional voxel grid by adopting a discretization volume reconstruction technology; establishing a geometric projection transformation model to reversely project the three-dimensional reconstruction volume to a two-dimensional projection domain, and minimizing the difference of the projection domain through joint optimization of Gaussian parameters; and introducing an adaptive density control mechanism to dynamically adjust Gaussian distribution. Compared with an existing method, the method has the advantages that an end-to-end discretization Gaussian function representation mode is adopted, global optimization of any CT geometric configuration is supported, high-quality rapid reconstruction within 5-15 minutes is achieved while the detail integrity of the anatomical structure is kept, and a high-precision and high-robustness three-dimensional reconstruction solution is provided for low-dose CT scanning and emergency treatment scenes.

The invention discloses an elevator shaft safety space data acquisition device and a three-dimensional reconstruction measurement method, an operator only needs to place the device in an elevator shaft, an elevator shaft three-dimensional data acquisition task can be automatically completed, the device can comprehensively cover each corner in the elevator shaft, the integrity and accuracy of data acquisition are ensured, and the safety space data acquisition efficiency is improved. High-quality original data is provided for spatial reconstruction and measurement; a data acquisition angle set by the method is based on an empirical value, and a foundation is laid for subsequent registration; according to the method, initial pose transformation and multi-view point cloud registration are combined, so that the stability of multi-view point cloud registration is ensured, and finally, redundancy of registered point cloud data is removed through combination of TSDF volume reconstruction. Operators do not need to enter and control in real time, three-dimensional data collection, reconstruction and measurement of the safe space of the elevator shaft are effectively achieved, operation is convenient, and efficiency is high.

A system and method combine optical and radiographic data to enhance imaging capabilities. Specifically, the system combines visually obtained patient pose position information and radiographic image information to facilitate calibrated surgical navigation. The process involves a data acquisition phase, a system calibration phase, a volume reconstruction phase, and a surgical navigation phase, all resulting in the alignment of instrument coordinates with the patient and reconstructed volume coordinates enabling tracking and navigation of surgical instruments within a reconstructed 3D volume of the patient anatomy, even if such anatomy is not exposed during a procedure.

The invention discloses a citrus X-ray image rapid reconstruction and defect segmentation method based on sparse point cloud and 3DGS, and relates to image processing, and the method comprises the following steps: S1, obtaining X-ray image data in a citrus under a sparse view angle; s2, performing three-dimensional reconstruction on the X-ray image data by adopting a three-dimensional Gaussian point cloud reconstruction method based on adaptive density control to obtain a reconstructed three-dimensional body; s3, performing automatic defect segmentation on the reconstructed three-dimensional body through the three-dimensional segmentation model to obtain two-dimensional defect mask slices and three-dimensional defect voxel data; and S4, according to the two-dimensional defect mask slices and the three-dimensional defect voxel data, performing comprehensive evaluation on the internal defects of the citrus, and outputting an internal quality report of the citrus. According to the method, full-process automatic processing from sparse view angle X-ray image data acquisition to three-dimensional volume reconstruction to automatic defect segmentation and type identification is realized, and the purpose of efficiently, accurately and losslessly identifying the internal defects of the citrus is achieved.

The invention relates to the technical field of radiation imaging, in particular to a two-step anisotropic Gaussian spattering method and device.The method comprises the steps that Gaussian kernel representation of a target scanning object is obtained through projection data initialization, and Gaussian kernel parameters are preliminarily optimized with a reconstruction result of a traditional reconstruction method as a supervision signal; a projection image represented by a Gaussian kernel is obtained through micro-rasterization, a scanned real projection image serves as a supervision signal, Gaussian kernel parameters are further optimized, anisotropic Gaussian regularization is introduced, the representation behavior of the Gaussian kernel is standardized, if a scanning mode does not meet a standard camera model, a virtual detection plane is introduced, and the real projection image is obtained. A mapping relation between the projection data and real projection data is constructed to meet a camera model; and finally voxelizing the Gaussian representation to obtain a computed tomography reconstruction result. According to the method, the volume reconstruction effect under the condition of CL scanning mode ill-conditioned data can be effectively improved, interlayer artifacts and noise of the reconstructed image are inhibited, and the practicability is extremely high.

A system and method combine optical and radiographic data to enhance imaging capabilities. Specifically, the system combines visually obtained patient pose position information and radiographic image information to facilitate calibrated surgical navigation. The process involves a data acquisition phase, a system calibration phase, a volume reconstruction phase, and a surgical navigation phase, all resulting in the alignment of instrument coordinates with the patient and reconstructed volume coordinates enabling tracking and navigation of surgical instruments within a reconstructed 3D volume of the patient anatomy, even if such anatomy is not exposed during a procedure.

The present invention discloses a method for identifying Mycobacterium tuberculosis in fluorescence microscopic images of sputum samples, comprising: performing multi-focal plane scanning on the sputum sample using an optical microscope to obtain a continuous fluorescence slice image sequence; processing the fluorescence slice image sequence using a three-dimensional volume reconstruction algorithm to generate three-dimensional fluorescence volume data containing spatial coordinates and fluorescence intensity information; converting the bacterial structure in the three-dimensional fluorescence volume data into a graph model, wherein nodes correspond to bacterial feature points and carry spatial coordinates and fluorescence parameters, and edges connect adjacent feature points and reflect spatial relationships; and performing feature analysis on the graph model using a graph neural network. In the present invention, high recognition accuracy is achieved, and the use of three-dimensional imaging and graph neural network technology can comprehensively obtain multi-dimensional information of the bacteria, construct a three-dimensional model and convert it into a graph model for feature analysis, effectively solving the problem of two-dimensional viewing angle limitation and greatly improving the accuracy of Mycobacterium tuberculosis identification.

The disclosure provides a three-dimensional CT imaging method and apparatus. The method includes: collecting two DR images with a perpendicular relationship for an object to be imaged; inputting the two DR images into a preset three-dimensional volume reconstruction model to obtain a three-dimensional volume of the object outputted by the three-dimensional volume reconstruction model; and slicing the three-dimensional volume to obtain three-dimensional CT imaging of the object, wherein the preset three-dimensional volume reconstruction model is trained using multiple training samples, each training sample comprising: a DR image group composed of two DR images with a perpendicular relationship, a pseudo-sinogram corresponding to the DR image group, and a three-dimensional volume corresponding to the DR image group.

A craniomaxillofacial three-dimensional reconstruction method based on free hand ultrasound comprises the following steps: (1) real-time volume reconstruction: tracking the position of an ultrasonic probe in real time by using a binocular camera, segmenting an ultrasonic image by using a double-flow neural network, and reconstructing a voxel grid in real time; (2) voxel denoising: eliminating outliers by using a morphological post-processing and template matching method, and then converting a voxel model into a point cloud; and (3) surface reconstruction: reconstructing the point cloud data into a polygonal mesh through a self-supervised implicit neural representation network to obtain a smooth bone surface.

The invention discloses a low-dose medical CT-oriented AI reconstruction and structure preserving denoising method, and relates to the technical field of AI. The method comprises the following steps: step 1, data preprocessing and initial feature map generation: obtaining a low-dose projection sequence and performing initial volume reconstruction, comparing the original projection with a forward projection obtained based on initial volume reconstruction calculation through a stripe consistency arbiter, and generating a stripe consistency label; step 2, performing iterative updating based on the stripe consistency label; and step 3, iterative optimization based on multi-party data comparison. According to the method, a unified index is established in an image domain and a projection domain, constraint on an important anatomical structure is realized by combining a structure anchoring graph and a tissue consistency field, and differentiated noise suppression and local credible signal backfilling are realized by utilizing strip consistency judgment and a ray sector reinjection mechanism, so that the ray dose is reduced, and the reliability of the system is improved. The marginal definition and tissue coherence are effectively maintained, and the diagnostic value and reliability of the low-dose CT image are remarkably improved.

A system and method combine optical and radiographic data to enhance imaging capabilities. Specifically, the system combines visually obtained patient pose position information and radiographic image information to facilitate calibrated surgical navigation. The process involves a data acquisition phase, a system calibration phase, a volume reconstruction phase, and a surgical navigation phase, all resulting in the alignment of instrument coordinates with the patient and reconstructed volume coordinates enabling tracking and navigation of surgical instruments within a reconstructed 3D volume of the patient anatomy, even if such anatomy is not exposed during a procedure.

The application provides a three-dimensional CT imaging method, device, electronic equipment and storage medium. The method comprises the following steps: collecting two DR images with a vertical relationship for an object to be imaged, and obtaining three-dimensional coordinate information of the object to be imaged; inputting the two DR images with the vertical relationship and the three-dimensional coordinate information into a preset three-dimensional volume reconstruction model to obtain three-dimensional volume data of the object to be imaged; and slicing the three-dimensional volume data to obtain a three-dimensional CT image of the object to be imaged. The method can improve the three-dimensional volume reconstruction speed on the basis of ensuring the imaging quality, and then quickly obtain a three-dimensional CT image with high quality.

The invention is entitled "System and method for local three-dimensional volume reconstruction using standard fluoroscope." A system for constructing fluoroscope-based three-dimensional volume data of a target region within a patient from two-dimensional fluoroscope images, including a marker structure, a fluoroscope imaging device configured to acquire a sequence of images of the target region and the marker structure, and a computing device. The computing device is configured to estimate a pose of the fluoroscope imaging device for at least a plurality of images of the sequence of images based on detection of a most likely and maximum probability of a projection of the marker structure as a whole on each of the plurality of images. The computing device is further configured to construct fluoroscope-based three-dimensional volume data of the target region based on the estimated pose of the fluoroscope imaging device.

The invention discloses a food calorie calculation method based on binocular vision, and relates to the technical field of computer vision and nutrition analysis. The method comprises the following steps: firstly, acquiring a food image by using a calibrated binocular camera, and preprocessing the image; and then a semi-global block matching algorithm is combined with weighted least square filtering to calculate and optimize a disparity map, and an edge-disparity collaborative network is introduced to improve the matching precision of the weak texture food. Then, segmenting a food area through a U-Net network, and outputting an accurate contour after edge optimization; and identifying the specific category of the food by using a lightweight classification network. And then, according to the food form, adaptively selecting an equivalent cylinder method or a point cloud reconstruction method to carry out three-dimensional volume reconstruction, and converting the volume into mass in combination with a density database. And finally, querying a calorie database containing cooking mode information to obtain an accurate calorie value. The method is obviously superior to the prior art, and an efficient and reliable solution is provided for family diet management and health monitoring.

The invention discloses a holographic three-dimensional bioassay method based on depth label coding and an attention mechanism. The invention relates to a holographic biometric platform employing depth label coding and end-to-end attention-enhanced three-dimensional reconstruction. The platform can enhance the depth of field and volume detection capability, and significantly improve the detection sensitivity and accuracy. A depth label coding strategy is integrated into a neural network architecture based on an attention mechanism, and three-dimensional volume reconstruction is converted into a dual-task framework combining bounding box detection and depth classification. And the limitation of a traditional generation type holographic reconstruction algorithm is overcome, so that accurate real-time three-dimensional positioning is realized. The holographic three-dimensional space positioning is beneficial to expanding field depth application and improving the density of the countable microspheres, chloramphenicol can be detected in a wide dynamic range, and compared with the traditional bioassay based on a two-dimensional hologram, the sensitivity is improved.

The invention discloses a collateral physiology intelligent evaluation system and method based on a coronary artery angiography image, and the system comprises an image collection module, a segmentation module, a reconstruction module, a modeling module, and an index calculation module. Wherein the image acquisition module is used for acquiring coronary artery angiography and screening minimum overlapped and shortened angiography projection images after CTO-PCI front and back side branch blood supply arteries and occluded blood vessels are opened; the segmentation module is used for carrying out frame-by-frame blood vessel segmentation on the obtained coronary artery angiography by adopting a deep learning algorithm; the reconstruction module is used for performing blood vessel three-dimensional volume reconstruction on the coronary artery binary image obtained through segmentation; the modeling module is used for modeling blood flow and calculating average blood flow; the index calculation module is used for calculating the collateral branch blood flow index according to the blood flow change before and after the CTO is opened. According to the method, the subjective defect of a traditional visual scoring system is eliminated, the operation process does not need to be interrupted, special consumables are not needed, and the medical cost and the occurrence risk of vascular injury related complications in the operation are effectively reduced.

The disclosure provides a three-dimensional CT imaging method and apparatus. The method includes: collecting two DR images with a perpendicular relationship for an object to be imaged; inputting the two DR images into a preset three-dimensional volume reconstruction model to obtain a three-dimensional volume of the object outputted by the three-dimensional volume reconstruction model; and slicing the three-dimensional volume to obtain three-dimensional CT imaging of the object, wherein the preset three-dimensional volume reconstruction model is trained using multiple training samples, each training sample comprising: a DR image group composed of two DR images with a perpendicular relationship, a pseudo-sinogram corresponding to the DR image group, and a three-dimensional volume corresponding to the DR image group.

The invention provides a semiconductor device detection method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring two DR images with a vertical relation for a semiconductor device to be detected; obtaining the three-dimensional volume of the semiconductor device to be detected corresponding to the two DR images with the vertical relation based on a preset three-dimensional volume reconstruction model; performing defect detection on the to-be-detected semiconductor device based on the three-dimensional volume; wherein the preset three-dimensional volume reconstruction model obtains the three-dimensional volume through large-scale structure recovery and small-scale detail reconstruction fusion. According to the method, the defect detection of the semiconductor device can be realized with high efficiency and high precision on the premise of saving manpower and material resources.

The present invention provides a method and system for self-supervised super-resolution reconstruction of magnetic resonance images based on a prior-guided diffusion model. The method comprises the following steps: image preprocessing, pseudo-pairing dataset construction, prior condition establishment, model training, model testing, and three-dimensional isotropic volume reconstruction. By constructing a self-supervised pseudo-pairing mechanism, the present invention significantly reduces reliance on true high-resolution images, enabling high-quality reconstruction of anisotropic MRI images into isotropic images. Simultaneously, by introducing multiple medical priors, including high-frequency residuals, edge structures, and region-of-interest guidance, the method effectively enhances image detail preservation and structural boundary clarity. While maintaining anatomical fidelity, it also improves model convergence speed and robustness, providing a technically feasible and clinically applicable solution for high-resolution reconstruction of medical images.

The invention relates to the field of 4DCT image data processing, in particular to a crack tracking model data preprocessing method based on a 4DCT image. Data volume reconstruction is performed based on coal rock sample synchronous loading and 4DCT scanning to obtain a coal rock sample dynamic fracture evolution sequence, angle deviation and background interference are eliminated by combining horizon slice integration, circular mask denoising and fracture rotation alignment, and the image contrast is enhanced by using gray threshold compression and fracture feature enhancement. And finally, accurately outputting the whole fracture spatio-temporal evolution process through a standardized preprocessing sample set input model. According to the method, the technical bottlenecks of discontinuous dynamic data, horizon offset and characteristic fuzziness of the traditional CT are overcome, the identification precision of the crack tracking model on microcrack initiation and expansion is remarkably improved, and a high-reliability dynamic observation method is provided for coal rock mechanical behavior analysis and engineering safety evaluation.

Methods and systems for generating animatable garments from a single image are disclosed. The system accesses a monocular image depicting a person wearing a fashion item, and generates a three-dimensional (3D) mesh representing the person in a normative space. The system determines a gesture of the person depicted in the monocular image and modifies the gesture of the 3D mesh to match the determined gesture of the person. The system extracts a portion of the 3D mesh corresponding to the fashion item, and generates an augmented reality (XR) item using the extracted portion of the 3D mesh.

The invention discloses a sputum sample fluorescence microscopic image mycobacterium tuberculosis identification method, which comprises the following steps: carrying out multi-focal plane scanning on a sputum sample by adopting an optical microscope to obtain a continuous fluorescence slice image sequence; processing the fluorescence slice image sequence through a three-dimensional volume reconstruction algorithm to generate three-dimensional fluorescence volume data containing space coordinates and fluorescence intensity information; a thallus structure in the three-dimensional fluorescence volume data is converted into a graph model, nodes correspond to thallus feature points and carry space coordinates and fluorescence parameters, and edges are connected with adjacent feature points and reflect a space relation; according to the method, the recognition precision is high, the three-dimensional imaging and graph neural network technology is adopted, the multi-dimensional information of the thalli can be comprehensively obtained, the three-dimensional model is constructed and converted into the graph model for feature analysis, the problem of limitation of a two-dimensional view angle is effectively solved, and the recognition accuracy is high. And the identification accuracy of the mycobacterium tuberculosis is greatly improved.

The invention discloses an artificial intelligence-driven three-dimensional digital volume-dependent deformation measurement system and method. The system comprises a three-dimensional speckle volume data acquisition unit, a three-dimensional coordinate normalization unit, a deep neural network prediction unit, a three-dimensional physical constraint unit and a three-dimensional field reconstruction output unit. The numerical stability of network input is ensured through three-dimensional coordinate normalization, synchronous prediction of a three-dimensional displacement field and a strain tensor field is realized by adopting an end-to-end neural network architecture, and physical rationality between the displacement field and the strain field is ensured by introducing volume reconstruction consistency constraint and three-dimensional deformation compatibility constraint. According to the invention, self-supervised learning is carried out only by relying on volume image pairs before and after deformation and physical constraints, synchronous and high-precision measurement of a three-dimensional displacement field and a strain tensor field is realized, and the problem of error amplification caused by three-dimensional numerical differentiation in a traditional method is avoided; the method is suitable for three-dimensional deformation measurement and analysis in the fields of material science, mechanical engineering, biomedical engineering and the like.

The application belongs to the technical field of image processing, and discloses a fast splicing method for CBCT vertical scanning, which comprises the following steps: step S1, a plurality of scanning positions of a rotating platform are arranged along the Z direction, and a scanning object is scanned at each scanning position to obtain scanning data; step S2, the angle offset of adjacent scanning positions is calculated according to the scanning data; volume reconstruction is performed on each scanning position; step S3, a common reconstructed slice image at the adjacent scanning positions is extracted, and the feature points of the slice image are determined by using a SIFT algorithm; the rotation angle of the slice image is determined by using a RANSAC algorithm; and step S4, the reconstructed volumes of the adjacent scanning positions are spliced after being rotated according to the rotation angle calculated in step S3. The application introduces a high-precision motion calibration technology and an improved image registration algorithm, the image quality is significantly improved, the joint seam artifact is eliminated, and a reliable guarantee is provided for high-resolution three-dimensional reconstruction.

The application discloses a shale oil and gas production dynamic prediction method based on GCN-GRU, and relates to the technical field of oil and gas well production prediction.According to the related parameters of each fractured horizontal well in a shale oil and gas well station, the volume reconstruction region spatial relationship and the artificial fracture distribution characteristics are combined, the interwell interference coefficient is calculated, the multi-well adjacency matrix is constructed, the production dynamic data is obtained according to the production of each fractured horizontal well in the shale oil and gas well station, the production dynamic data is distributed to a training set and a test set, the GCN-GRU model is constructed, the production of the fractured horizontal well is predicted by using the production dynamic data in the training set and the multi-well adjacency matrix to train the model, and the accuracy of the model prediction after training is verified by using the production dynamic data in the test set.The interwell interference is quantified, and the high-precision prediction of the shale oil and gas production under the conditions of three-dimensional development and large-scale hydraulic fracturing is realized.

Slice alignment based on reconstructed volume. Disclosed herein are apparatuses and methods for aligning a charged particle beam based on a volume reconstruction. An example method includes at least: forming a reconstructed volume of a portion of a sample, the sample including a plurality of structures, and the reconstructed volume including a portion of the plurality of structures; performing a mathematical transformation on each of a plurality of planes of the reconstructed volume over a range of angles; and determining a target orientation of the sample over the range of angles based on the mathematical transformation of each of the plurality of planes, wherein the target orientation aligns the plurality of structures parallel to an optical axis of a charged particle beam.

The invention provides an ultrafast multi-dimensional super-resolution imaging method based on fluctuation high-order acoustic reconstruction, belongs to the technical field of super-resolution imaging, and solves the problem that in the existing ultrasonic imaging technology, when the imaging frame rate is reduced, the imaging effect and the imaging speed are obviously reduced. Comprising the following steps: S1, performing singular value decomposition filtering in a time domain to extract a microbubble signal; s2, extracting MB flow direction information by using MB direction filtering in the Fourier domain; s3, background filtering is carried out in the wavelet domain; s4, interpolation is carried out in a Fourier domain, and then a Richardson-Lucy (RL) pre-deconvolution step is carried out; s5, further separating the activity of the MB by using MB density filtering in the time domain; s6, performing high-order accumulation calculation; s7, using a three-dimensional sparse convolution algorithm to improve the spatial resolution and the imaging quality; and S8, completing volume reconstruction by using data fusion.

The present invention discloses an elevator shaft safety space data acquisition device and a three-dimensional reconstruction measurement method. The operator only needs to place the device in the elevator shaft to automatically complete the three-dimensional data acquisition task of the elevator shaft. The device can fully cover all corners in the elevator shaft, ensuring the integrity and accuracy of data acquisition, and providing high-quality original data for space reconstruction and measurement. The data acquisition angle set by the method is based on empirical values, laying a foundation for subsequent registration. By combining the initial pose transformation and multi-view point cloud registration, the stability of multi-view point cloud registration is ensured. Finally, through the combination of TSDF volume reconstruction, the redundancy of the registered point cloud data is removed. There is no need for the operator to enter and perform real-time control, effectively realizing the three-dimensional data acquisition, reconstruction and measurement of the elevator shaft safety space, with convenient operation and high efficiency.

A system and method combine optical and radiographic data to enhance imaging capabilities. Specifically, the system combines visually obtained patient pose position information and radiographic image information to facilitate calibrated surgical navigation. The process involves a data acquisition phase, a system calibration phase, a volume reconstruction phase, and a surgical navigation phase, all resulting in the alignment of instrument coordinates with the patient and reconstructed volume coordinates enabling tracking and navigation of surgical instruments within a reconstructed 3D volume of the patient anatomy, even if such anatomy is not exposed during a procedure.