12 results about "Morphological model" patented technology

The application provides a virtual organ simulation method and system based on three-dimensional stacked spatial transcriptome data, which comprises the following steps: constructing a spatial adjacency network according to the coordinate information of a plurality of spatial transcriptome slice data; using a graph attention autoencoder to perform expression embedding representation learning on each point in the spatial adjacency network; voxelizing the three-dimensional coordinates of each slice, constructing a continuum through Gaussian smoothing, and generating a three-dimensional morphological model; training a latent space diffusion model to generate a latent expression of the spatial transcriptome data; for a given coordinate and condition, sampling an embedding vector from the trained latent space diffusion model, and inputting the embedding vector into the graph attention autoencoder to obtain the gene expression profile of the corresponding point. The advantage of the application lies in that the spatial transcriptome of the whole organ or even the cross-tissue region is uniformly modeled and expressed, which breaks through the small-scale modeling capability of the existing method which is limited to a local region or a single slice.

The invention provides a lingual surface three-dimensional shape reconstruction method and system based on deep learning, and relates to the technical field of three-dimensional reconstruction, and the method comprises the steps: activating an image collection array based on a light source illumination parameter to carry out the multi-angle two-dimensional sensing of a target tongue part, and obtaining an original lingual surface image sequence; performing visual cutting, extracting a plurality of lingual surface target images, performing deep learning, and constructing a lingual surface depth map; performing three-dimensional shape analysis, constructing lingual surface three-dimensional point cloud coordinates, mapping the coordinates to a plurality of lingual surface target images for identification, and performing comparative analysis on an identification result and a standard image library to generate a comparative parameter set; and performing correction, performing three-dimensional reconstruction on the target tongue according to a correction result, and constructing a three-dimensional morphological model. The technical problems that a lingual surface analysis method in the prior art mostly depends on two-dimensional images or manual observation, subjectivity is often high, information loss is serious, and the real three-dimensional form of the lingual surface is difficult to obtain are solved.

This invention discloses an omnidirectional morphological scanning device suitable for large-size rock specimens, comprising a support platform, a hydraulic lifting system, a scanning system, and a computer system. The hydraulic lifting system includes a first set of lifting devices and a second set of lifting devices. The scanning system includes scanning devices mounted on the support platform and arranged around the rock specimen, and scanning devices arranged to the sides of the rock specimen. Each scanning device includes a hyperspectral camera, a high-definition camera, and a vertical angle adjustment device. The computer system controls the first and second sets of lifting devices to simultaneously or alternately lift the rock specimen. The computer system receives data captured by the hyperspectral camera, matches it with a built-in mineral spectral library to determine mineral types, and analyzes the mineral composition of the rock specimen surface. Data captured by the high-definition camera is preprocessed and aligned with multiple perspectives to complete model reconstruction and analyze the morphology of the rock specimen surface. This invention optimizes the structure of the morphological scanning device and improves the accuracy of morphological model stitching.

The present disclosure relates to the technical field of three-dimensional model construction, and provides a method and apparatus for reconstructing a three-dimensional model, which may solve the problems of the structure being complex and modeling efficiency being low for existing three-dimensional models. The method for reconstructing a three-dimensional model of the present disclosure comprises: according to data features of a group of initial three-dimensional models, acquiring point cloud data of each initial model; selecting one model in the group of initial three-dimensional models as a standard three-dimensional model, and acquiring a plurality of discrete deformation fields by using difference values between points in the standard three-dimensional model and points in the other initial three-dimensional models; converting the plurality of discrete deformation fields into a statistical morphological model by using a multi-dimensional Gaussian process; and resampling the statistical morphological model, and acquiring a target three-dimensional model.

This invention discloses a method for quantitatively evaluating the morphology of aggregates based on frequency analysis, comprising the following steps: performing three-dimensional laser scanning on aggregate particles and removing noise to obtain the boundary pixel coordinates of the aggregate particles; calculating the centroid coordinates and centered coordinates of the aggregate particles' outlines, and constructing a complex sequence of each pixel point in the aggregate particles based on the centered coordinates; performing a discrete Fourier transform on the complex sequence to calculate the Fourier coefficients and amplitude spectrum of the complex sequence; extracting feature values ​​of different morphological parameters according to a preset set of contribution frequencies, and constructing a morphological model of each morphological parameter based on the feature values; substituting the amplitude spectrum of the aggregate particles into the morphological model to quantify each morphological parameter of the aggregate particles. This invention also discloses a system for quantitatively evaluating the morphology of aggregates based on frequency analysis. This invention solves the technical problem of how to accurately and quickly determine the morphology of aggregate particles.

The invention discloses a novel anti-bias-fixation chip preparation method, and relates to the technical field of semiconductor chip preparation, and the method comprises the steps: generating a growth column target morphological model based on target geometric parameters; preparing a growth column in the electrode area; comparing the three-dimensional contour information of the growth column with the target morphological model of the growth column to generate a morphological error matrix; correcting the etching track by using the morphological error matrix, and constructing an infiltration energy level gradient which is gradually increased from the center to the edge at the top end of the growth column; calculating thermodynamic stability values of the growth columns based on the infiltration energy level gradient, screening out qualified growth columns based on the thermodynamic stability values, and constructing a growth column array; and based on the growth column array, generating a self-alignment liquid bridge restoring force in a solid crystal melting stage, and inducing the chip to perform self-alignment posture correction to obtain the flip LED chip. According to the invention, active pose correction and high-precision anti-deviation preparation of the flip LED chip are realized.

The invention belongs to the field of oil and gas exploitation, particularly relates to a casing deformation prediction method, system and equipment, and aims to solve the problem that a method for predicting casing deformation caused by sand production is not scientific and accurate enough. The method comprises the following steps: determining a sand production cavity form model of an oil and gas well and parameters of the sand production cavity form model; based on the parameters, the sand production condition and the perforation condition of the oil and gas well are combined, and the axial pressure borne by the sleeve at the sand production cavity of the oil and gas well is calculated through a sleeve stress geometry and mathematical model and a sand production cavity volume calculation formula of the oil and gas well; determining the residual wall thickness of the sleeve; calculating the critical stress of the casing according to the type and the residual wall thickness of the casing; calculating the deformation critical force of the casing pipe of the oil and gas well based on the critical stress in combination with the compression area of the casing pipe; if the axial pressure borne by the casing pipe is larger than the deformation critical force, the casing pipe deforms. According to the invention, accurate prediction of casing deformation is realized.

The invention relates to the technical field of remote monitoring, and discloses a management method and system applied to protection and rejuvenation of ancient and famous trees, and the method comprises the steps: generating basic information, physiological and biochemical indexes, growth environment parameters, and dynamic baseline data of pest and disease damage and safety inspection data; establishing a three-dimensional shape model of the ancient and famous trees to analyze the growth stages of the ancient and famous trees; analyzing the ancient tree health index of the ancient and famous tree in the growth stage by using a preset ancient tree health analysis model; analyzing health key influence factors of the ancient and famous trees, and when the health index of the ancient trees is lower than a preset health threshold value of the ancient trees, establishing an etiological traceability map of the ancient and famous trees through the key influence factors; and analyzing a personalized rejuvenation path of the ancient and famous trees through the etiological traceability map to generate a rejuvenation task work order of the ancient and famous trees, and performing protection and rejuvenation of the ancient and famous trees based on the rejuvenation task work order. The scientific, refined and sustainable management efficiency of large-scale ancient tree groups can be improved.

The application discloses a three-dimensional dynamic biological simulation method based on formula modeling, and the method comprises the following steps: initializing a three-dimensional simulation space and initial positions, velocities and morphological parameters of a biological body, and constructing a parameterized biological morphological model; dynamically updating a two-dimensional morphological reference of the biological body and positions and postures of the biological body in a three-dimensional space based on morphological modulation parameters and motion velocities; for different structural parts of the biological body, an equal-angle slicing or random-direction generation equal-difference strategy is adopted to convert the two-dimensional morphological reference into a three-dimensional entity structure with symmetry or randomness, and linear interpolation is performed to enhance continuity; finally, the three-dimensional entity structure is rendered into volume data, superimposed intensity projection is performed along a preset direction, continuous frame images are generated, and the continuous frame images are output as multi-page TIFF format simulation data after a preset number of frames is reached.

The invention provides a unicondylar replacement guiding and positioning equipment design method and related equipment. The design method comprises the following steps: acquiring historical medical data of a unicondylar osteotomy object; generating a knee joint statistical morphology model based on the historical medical data; and generating a predetermined unicompartmental knee arthroplasty guiding and positioning device based on the knee joint statistical morphology model. According to the method, the knee joint statistical morphology model is generated through the historical medical data, and then the unicompartmental knee arthroplasty guiding and positioning equipment is generated based on the statistical morphology model, so that the generated unicompartmental knee arthroplasty guiding and positioning equipment is matched with an individual, and the osteotomy precision and the operation effect of unicompartmental knee arthroplasty are improved.

This invention discloses a method, device, storage medium, and apparatus for tree identification using a drone. The invention preprocesses remote sensing image information acquired by the drone, segments the remote sensing tree images based on an improved Mask-R-CNN model, and identifies tree species based on a preset tree species model and tree image information, obtaining image sets corresponding to each tree species type. Based on a preset morphological model and the image sets of each tree species type, the invention predicts the growth trend of trees within a target area, and determines the perceived growth status of trees in the target area based on the prediction results. This invention segments the preprocessed remote sensing tree images using an improved Mask-R-CNN model and identifies tree species based on a preset tree species model, thereby determining the growth status of each type of tree. This invention achieves accurate vegetation identification and precise prediction of tree growth status using a drone, improving data accuracy.

This invention relates to the field of artificial intelligence technology and discloses a method and system for real-time 3D paleontological reconstruction based on artificial intelligence and multimodal interaction. It aims to solve the problems of fragmented multi-source data, low automation, limited interaction methods, and lack of intelligent reasoning capabilities in existing technologies, which lead to strong subjectivity and difficulty in dynamic verification. The method includes: constructing a four-dimensional unified representation framework integrating fossil point clouds, stratigraphy, isotopes, and extant analog data, and generating a structured knowledge graph; using conditional generative adversarial networks to complete missing structures under biomechanical and evolutionary rules constraints, generating an anatomically consistent 3D morphological model; deploying natural language and virtual reality interfaces to achieve multimodal human-computer collaborative interaction; and triggering incremental learning based on user feedback and new evidence to complete real-time iterative optimization. This system integrates the above modules and scientific verification mechanisms, enabling high-fidelity, interactive, verifiable, and continuously evolving 3D dynamic paleontological reconstruction.