635 results about "3D reconstruction" patented technology

The invention discloses a three-dimensional scene reconstruction method and device based on large model geometric prior, and a medium, and aims to solve the problems that a conventional 3DGS is liable to have artifacts and detail loss in geometric discontinuity, data redundancy and illumination variation scenes, and predicts a dense depth map and a normal map from a monocular image by using a pre-trained large model. The position and form of the Gaussian kernel are constrained as additional geometric priori; a primitive adjustment strategy based on kernel density estimation is introduced in the training stage, small Gaussian primitives with similar structures and adjacent spaces are combined into a large Gaussian primitive, the rendering quality is kept, redundancy is reduced, and the volume of the model is reduced; an exposure coefficient is adaptively estimated for each input image, an exposure compensation image loss function is constructed, and floating artifacts caused by illumination differences at shooting moments are eliminated. Experiments show that compared with the prior art, the method improves the three-dimensional reconstruction precision and real-time rendering quality of complex illumination and less-texture areas in a public data set and an unmanned aerial vehicle aerial photography scene.

The invention discloses a single-view large-scale outdoor scene three-dimensional reconstruction method based on three-dimensional Gaussian splashing, and the method comprises the steps: collecting a pseudo aerial image, and constructing a panoramic multi-mode supervision end-to-end single-view three-dimensional reconstruction model; meanwhile, panoramic consistency supervision, semantic constraint depth regularization and a radial weighted luminosity loss and Gaussian cutting mechanism are introduced, so that the defect of insufficient geometric constraint of traditional single-view three-dimensional reconstruction is effectively overcome, and high-efficiency and high-fidelity three-dimensional modeling of a large-scale outdoor scene under single image input is realized; the method is suitable for various actual scenes such as smart city construction, automatic driving simulation, virtual reality / augmented reality, digital twinning and the like.

The invention discloses an immovable cultural relic three-dimensional reconstruction system which comprises a data acquisition module, a feature extraction module, a restoration simulation module and an optimization and decision module. The data acquisition module is used for acquiring three-dimensional data of an antique; the feature extraction module is used for extracting key features of antiques by using a deep learning algorithm; the restoration simulation module is used for constructing a virtual restoration model based on the extracted features, simulating an antique restoration process and evaluating a restoration scheme; and the optimization and decision module is used for realizing precise positioning and navigation of a repair site in combination with the SLAM technology, optimizing a repair path and assisting a repairer in making a scientific decision. According to the method, the repairing efficiency and precision are remarkably improved, automation and intelligentization of the ancient object repairing process are achieved by introducing advanced technologies such as deep learning and NeRF, the repairing efficiency and precision are greatly improved, and the repairing period is shortened.

The invention relates to the technical field of three-dimensional reconstruction, in particular to a near-infrared assisted low-light scene three-dimensional reconstruction method based on 3D Gaussian splashing, which is characterized in that generation of a Gaussian ellipsoid is dominated by a near-infrared image with a high signal-to-noise ratio, and a stable geometric basis is provided for a visible light information missing area; in the rendering stage, a near-infrared rendering image and a normal visible light rendering image are respectively generated through Gaussian ellipsoid shared geometric parameters and respective opacity and color attributes of two modes; through cross-modal structure similarity loss, a normal visible light image which is forcibly rendered is aligned with a near-infrared light image in structure, and clear near-infrared structure information is used to strictly constrain the recovery process of the color of the visible light image, so that the accuracy and authenticity of a recovery result are ensured.

The invention discloses a three-dimensional reconstruction method based on a pulse camera and electronic equipment, and relates to the technical field of three-dimensional reconstruction. According to the method, the problems of insufficient spatial-temporal information modeling and inaccurate initial attitude estimation in pulse camera three-dimensional reconstruction in related technologies are effectively solved, the precision of a three-dimensional reconstruction model is improved, a three-dimensional scene with higher quality and more realistic sense can be generated, meanwhile, efficient alignment of pulse data and three-dimensional geometry is realized through joint optimization of a framework, and the accuracy of the three-dimensional reconstruction model is improved. And the overall robustness and generalization ability of the system are enhanced.

A method and system for constructing a three-dimensional (3D) aerial survey of a city street scene include obtaining a plurality of video frames from a calibrated multi-camera setup covering a 360-degree view mounted on a moving vehicle. The plurality of video frames is split into a plurality of 3D parts containing a subset of the plurality of video frames and preprocessing the subset of the plurality of video frames of each part of the plurality of parts to obtain a calculated information. Further, constructing, by the processing circuitry, a 3D representation of each part of the plurality of parts based on the calculated information to obtain a plurality of local 3D reconstructed scene intervals. The method includes stitching and filtering, by the processing circuitry, the plurality of local 3D reconstructed scene intervals to construct the 3D city street scene.

The invention discloses a video gait analysis-based asthenia syndrome evaluation method, medium and equipment, and the method comprises the steps: firstly, synchronously collecting gait videos of different visual angles through a plurality of cameras, and extracting 2D joint point sequences of all visual angles through a stacked hourglass network; then, a high-precision 3D joint point time sequence is reconstructed through a multi-view fusion algorithm, depth features are extracted through a residual connection network, and data of all view angles are dynamically weighted and fused based on the shielding rate and the confidence coefficient; performing sliding window segmentation and downsampling on the 3D sequence, and performing action classification by using an LSTM time sequence model; and finally, key motion parameters are extracted from a classification result to calculate a weakness score, and weakness grade evaluation is realized. According to the method, the problem of single-view shielding is solved through multi-view 3D reconstruction, the evaluation precision is remarkably improved by combining dynamic weighted fusion and time sequence modeling, and compared with a traditional method, the method has the advantages of being non-contact, low in cost and high in objectivity, and is suitable for large-scale screening of community and family environments.

The invention discloses a self-adaptive sensing anti-shake regulation and control method and system for a spinal surgery auxiliary mechanical arm, and relates to the technical field of spinal surgery, and the method comprises the steps: collecting multi-dimensional data of a mechanical arm tail end actuator through a multi-dimensional sensor group, and transmitting the multi-dimensional data to an AeroStead space-level stability calculation platform; the platform carries out filtering processing on the data, and vibration of the mechanical arm and disturbance components caused by external interference are separated; constructing a spine mechanical arm rigid-flexible coupling dynamic disturbance observation and compensation model by using the disturbance component to observe disturbance and generate a compensation amount; calling a spinal operation area three-dimensional reconstruction and dynamic obstacle avoidance mixed density network model, and fusing different images to generate an operation area three-dimensional dynamic model and an obstacle avoidance adjustment signal; a mechanical arm motion control instruction is generated through a force sense position cooperation self-adaptive control algorithm; and the mechanical arm is driven to move for closed-loop regulation and control. According to the method, through fusion of multiple models and algorithms, the disturbance compensation precision, the operation area modeling accuracy and the control collaboration are improved, and the operation safety is guaranteed.

The invention discloses a sparse visual angle three-dimensional reconstruction method and system based on voxel grid constraint, and belongs to the technical field of visual three-dimensional reconstruction. Constructing a three-dimensional voxel grid of a self-adaptive scene scale based on point cloud distribution, and dividing the point cloud to the corresponding voxel grid; fusing the geometric features of the fast point feature histogram and the confidence score in the grid, generating a geometric confidence comprehensive measure, and screening key points to initialize Gaussian primitives; designing a voxel grid constrained gradient clipping strategy, limiting Gaussian primitive error diffusion through a distance attenuation coefficient, and adaptively optimizing grid distribution in combination with a dynamic grid deletion and addition mechanism; and finally, carrying out iterative training by using a 3D Gaussian splash radiation field loss function to realize high-fidelity static scene reconstruction under a sparse view angle. According to the method, scene geometric priori is introduced, an optimization strategy based on voxel grid constraint is designed to effectively control excessive diffusion or drift of Gaussian primitives, generation of artifacts is reduced, and meanwhile, the situation that robustness is reduced due to the influence of priori quality is avoided.

The invention relates to the field of virtual reality, in particular to an adaptive virtual reality cognitive training method and system. The method comprises the steps of obtaining a user individual information set and a real-time multi-modal physiological data stream, performing real environment three-dimensional reconstruction and relative and friend face modeling based on the user individual information set, and generating an individual virtual training scene; based on the real-time multi-modal physiological data flow and the individualized virtual training scene, nerve-behavior feature extraction and digital twinning dynamic modeling are carried out, and a user cognitive state evaluation information set is generated; based on the user cognitive state evaluation information set, training task parameter real-time optimization and virtual scene element dynamic adjustment are carried out through an adaptive strategy, and a cognitive enhancement intervention instruction set is generated; and based on the cognitive enhancement intervention instruction set, performing neural regulation intervention and training data closed-loop feedback, and generating and outputting an individualized cognitive training report. According to the method and the device, accurate self-adaptive intervention is realized in a virtual reality cognitive training process.

The invention discloses a three-dimensional visualization head and neck tumor preoperative imaging system based on deep learning. The method comprises the following steps: image analysis and preprocessing; carrying out image denoising and spatial-temporal feature extraction based on a deep learning convolutional neural network architecture, and introducing a u-Net model to realize image segmentation and semantic annotation; performing three-dimensional reconstruction and surface rendering by adopting a Marking Cubes algorithm, and realizing visual interaction image presentation and personalized operation by combining volume rendering and surface rendering; and data is output through multiple ports, and data viewing and interactive operation of a web terminal and a mobile terminal are supported. The method has the advantages that according to the technical scheme, through fusion of a three-dimensional attention mechanism and multi-phase data, the tumor segmentation Dice coefficient reaches 0.92 + / -0.03, and the method is superior to a traditional threshold segmentation method.

The invention relates to the field of low-altitude economy and geographic information, in particular to a low-altitude scene-based twinborn data rapid updating method, which comprises the following steps of: acquiring multi-source sensing data of a target low-altitude area in real time; comparing with a historical twinborn model, and identifying a change area and evaluating a priority by using an artificial intelligence algorithm; dynamically generating an optimal unmanned aerial vehicle data acquisition route by combining the airspace information and priority of the change area; then the unmanned aerial vehicle is controlled to collect updated data along the route; performing lightweight real-time three-dimensional reconstruction on the updated data to form a local updated model; and finally, fusing the model with a historical twinborn model to generate a final target twinborn model. Through intelligent change identification, dynamic route planning, lightweight reconstruction and incremental updating, the updating efficiency is significantly improved, the resource consumption is reduced, the real-time accuracy of the model is ensured, and the method is suitable for multiple fields such as low-altitude economy and urban governance.

A computer-implemented method for 3D reconstruction of a target is provided, comprising obtaining an initial global reconstruction of the target in a 3D space, inferred by a global machine learning model; providing, to a user, an initial visualisation of the target based on the reconstruction; receiving, from the user, at least one indication of at least one point of interest in the visualisation; resampling at least one first subsection of the target based on the at least one point of interest to obtain local data, wherein the local data is associated with the subsection based on spatial information that associates the local data with a point in 3D space; inputting the resampled local data and spatial information into a local feature machine learning model to obtain at least one 3D reconstruction of the target, wherein the local feature machine learning model has been trained to output a target reconstruction from local data of resampled subsections, and wherein the 3D coordinate system of the local 3D reconstruction aligns with the global 3D reconstruction; and merging the global 3D reconstruction with the local 3D reconstruction. A corresponding computer system and computer readable medium may also be provided.

The invention belongs to the technical field of computer vision and three-dimensional reconstruction, and particularly relates to a building intelligent three-dimensional reconstruction method and system based on point cloud, and the method comprises the steps: obtaining the point cloud data of a building, and carrying out the preprocessing of the point cloud data; constructing an intelligent three-dimensional reconstruction model based on an intelligent algorithm selection mechanism, an adaptive surface reconstruction technology and an anti-mold-penetration processing algorithm; and three-dimensional reconstruction of the building is realized based on the preprocessed point cloud data and the intelligent three-dimensional reconstruction model. Through an intelligent algorithm selection mechanism, a self-adaptive surface reconstruction technology and an anti-mold-penetration processing algorithm, the technical problems of blind algorithm selection, unstable reconstruction quality, geometric mold penetration and the like in the prior art are solved, and high-quality and high-efficiency building three-dimensional reconstruction of point cloud data of various formats is realized.

A comprehensive spatial AI platform for neural 3D reconstruction and built environment analysis integrates foundation models, deep learning methods, and spatial reasoning capabilities to provide expert knowledge of the physical world. The platform combines symbolic AI and machine learning to facilitate 3D semantics for insurance, real estate, construction, robotics, and other business applications. The system processes captured images, videos, or point clouds through neural networks with low compute requirements, incorporating device-agnostic advanced spatial intelligence that delivers geometric, semantic, and relational data. The platform includes proprietary training innovations, comprehensive measurement and semantic understanding, external sensor integration, human-in-the-loop quality assurance, and API integration for programmatic access. Advanced spatial reasoning capabilities enable property damage assessments, construction progress tracking, robotics navigation, and real-time applications including room dimension validation and automated repair estimates, supporting enterprise workflows across various industry segments while enabling productivity improvements and new value-added spatial AI use cases.

The invention relates to a laser radar point cloud densification method and device fusing image information, and a medium. The method comprises the following steps: fixing a visual camera and a laser radar on a rigid tool, and carrying out joint calibration and data alignment; a laser radar is used to collect three-dimensional point cloud data, and a visual camera is used to shoot a scene; performing super-pixel segmentation on the image by using an image brightness linear iterative clustering algorithm; mapping the three-dimensional point cloud data into a segmented two-dimensional image superpixel pattern spot region by using a jointly calibrated camera model conversion relationship to realize feature clustering of the original three-dimensional point cloud of the laser radar; performing curved surface fitting on a clustered result by using a random sampling consistency algorithm; and linear interpolation is carried out on the fitted curved surface, dense points which are not covered by the original point cloud of the laser radar are supplemented and generated, and densification of the point cloud of the laser radar is realized. According to the method, the point cloud densification precision and practicability are improved, and technical support is provided for automatic driving, robot navigation and three-dimensional reconstruction.

The invention discloses a miniature intelligent terminal intestinal image processing method and device, a storage medium and computer equipment. The method comprises the following steps: preprocessing an original intestinal tract image set acquired by a monocular endoscope, calculating according to the preprocessed intestinal tract image set to obtain a motion calibration scale factor and a texture calibration scale factor, and adaptively fusing to obtain a conversion coefficient from a pixel to an actual size; extracting polyp region key feature points of each image of the preprocessed intestinal image set, constructing a Delaunay triangulation graph according to each polyp region key feature point, and performing non-rigid deformation matching based on a cross-graph convolutional network and an optimal transmission algorithm to obtain a matched feature point set; and sparse 3D reconstruction is carried out according to the matched feature point set to obtain a polyp point cloud, and the polyp size is calculated according to the Euclidean distance between two farthest points in the polyp point cloud and the conversion coefficient. According to the method, the problems of monocular image scale ambiguity and non-rigid deformation can be solved, and the more accurate polyp size can be obtained.

A mobile robot system is described in having a mobile robot and cloud system. The mobile robot leverages cloud computing to offload Neural Radiance Fields (NeRF) based 3D scene reconstruction. The mobile robot advantageously adopts techniques for view filtering and next-best view selection that optimize the image collection process necessary for training an NeRF model with the cloud system. These techniques enable the mobile robot to discard redundant images that do not provide significant new information about the environment. Additionally, these techniques enable the mobile robot to strategically select next-best views that maximize the information gain, while minimizing a total number of images required and the time required to capture the images. These techniques provide a significant reduction in the overall bandwidth required for providing image data to the cloud system and can result in a more accurate and higher quality 3D reconstruction of the environment.

Disclosed is a method for enabling exposure-guided three-dimensional (3D) reconstruction, including (i) receiving Low Dynamic Range (LDR) input image captured by camera; (ii) rendering High Dynamic Range (HDR) image corresponding to input image, using HDR 3D reconstruction model pre-trained for HDR image reconstruction of 3D environment; (iii) applying HDR to LDR tone-mapping operator on HDR image, for producing tone-mapped LDR image; (iv) determining first loss function (FLF) between input image and tone-mapped LDR image, FLF comprising weighted pixel value differences between pixels of input image and corresponding pixels of tone-mapped LDR image; (v) determining whether input image has saturated pixel(s), saturated pixel(s) comprises: highlight saturated pixel, shadow saturated pixel; (vi) de-weighting pixel value difference corresponding to saturated pixel in FLF; (vii) back-propagating gradient of FLF with respect to model parameters, through differentiable render function of HDR 3D reconstruction model, for adjusting model parameters in way that reduces FLF.

The invention discloses a light field microscopic three-dimensional imaging method combining self-supervised learning and optical constraint. The method comprises the following steps: constructing a light field microscopic three-dimensional reconstruction reference data set; according to the invention, the two-dimensional light field image collected by the microscope is divided into a plurality of sub-aperture views, angle-space joint representation is formed, angle information and space information in the light field image are captured more comprehensively, and a richer and more accurate feature basis is provided for subsequent three-dimensional reconstruction; a self-supervised pre-training mode is adopted, so that the model has a certain knowledge basis in an initial stage, and in subsequent supervised training or fine tuning, convergence can be faster, and the training efficiency is improved; according to the method, physical constraints are integrated into the training process, so that the reconstructed three-dimensional structure is not only visual reasonable, but also credible in physical significance, the designed loss function can more accurately measure the difference between the reconstruction result and the real three-dimensional structure, and the quality and precision of the reconstruction result are improved.

The invention provides an endoscope image real-time dynamic three-dimensional reconstruction method and system, and the method comprises the steps: obtaining initial monocular endoscope image data, constructing an endoscope image training data set through a Gaussian splash algorithm and a chaos algorithm, constructing a 3D reconstruction constraint network, and carrying out the real-time dynamic three-dimensional reconstruction of an endoscope image. The 3D reconstruction constraint network is pre-trained in combination with the endoscope image training data set and a preset physical loss function, real-time monocular endoscope image data is obtained, a dynamic 3D Gaussian scene is generated, scene rendering is conducted on the dynamic 3D Gaussian scene of the endoscope based on endoscope visual parameters corresponding to the real-time monocular endoscope image data, and the dynamic 3D Gaussian scene of the endoscope is obtained. The method comprises the steps of generating ideal monocular endoscope image data, obtaining a discrimination update parameter group by adopting a rendering discrimination update mechanism, and performing feedback optimization on a 3D reconstruction constraint network based on the discrimination update parameter group. The accuracy, the speed and the robustness of three-dimensional reconstruction of the endoscope are improved.

The invention discloses an aero-engine single crystal turbine blade coating thickness uniformity detection method based on AI and three-dimensional reconstruction, and relates to the technical field of aero-engine component detection. According to the method, structural data of the turbine blade before coating and after coating are collected and subjected to three-dimensional reconstruction, double-model accurate alignment and differential processing are achieved through an AI registration model, and a coating three-dimensional model is extracted; and through voxelization processing and an AI thickness analysis model, obtaining thickness data of a whole region and a key region, carrying out multi-dimensional evaluation in combination with a uniformity coefficient, and finally judging the qualification of the coating and outputting a comprehensive report. According to the method, global accurate detection of the coating thickness and key analysis of key areas can be realized, the detection efficiency and accuracy are remarkably improved, and the high-precision quality control requirement of the aero-engine turbine blade coating is met.

The invention relates to the field of three-dimensional scene reconstruction, in particular to a three-dimensional Gaussian reconstruction method based on unbiased depth and normal supervision, and the method comprises the steps: obtaining a monocular estimation depth map of an image based on a monocular depth estimation model, and carrying out the correction through employing a sparse three-dimensional point cloud calibration alignment method; constructing a scale constraint loss function, and constraining the three-dimensional Gaussian ellipsoid into a two-dimensional Gaussian plane along the surface normal direction of the three-dimensional Gaussian ellipsoid; obtaining a rendering depth map based on the surface normal direction of the two-dimensional Gaussian plane and the sight line direction of the camera; constructing a depth loss function according to the rendering depth map and the calibration monocular estimation depth map; calculating an estimation surface normal of each pixel of the calibration monocular estimation depth map, and calculating a direction consistency loss function in combination with the rendering normal map; the scale constraint loss function, the depth loss function and the direction consistency loss function are weighted to obtain overall optimization loss, the overall optimization loss is subjected to back propagation, and finally a three-dimensional reconstruction result is obtained; according to the method, the geometric accuracy of three-dimensional Gaussian reconstruction is improved.

The invention relates to a dynamic object three-dimensional reconstruction method and device, computer equipment and a storage medium. The method comprises the following steps: extracting a cross-frame point cloud sequence of a target object from historical automatic driving data; processing the cross-frame point cloud sequence through space-time alignment and motion trail compensation to obtain a high-density accumulated point cloud; generating rays according to the sensor pose data and the high-density accumulated point cloud, sampling along the rays, and generating training data based on a sampling result; a double-branch neural network comprising geometric branches and radiation field branches is constructed, the geometric branches are used for modeling an object surface by adopting a symbolic distance field, and the radiation field branches are used for learning point cloud density and surface attributes by adopting a neural radiation field; performing iterative training on the double-branch neural network based on the training data; and extracting network parameters of the trained double-branch neural network to obtain a dynamic implicit representation model of the target object. According to the invention, the three-dimensional reconstruction precision of the dynamic object can be improved, and the space-time information storage efficiency of the dynamic object is improved.

The invention provides a plant three-dimensional reconstruction method and system based on image segmentation and projection guidance, and is applied to the technical field of plant three-dimensional reconstruction, and the method comprises the steps: obtaining a multi-view image sequence of a target plant, the multi-view image sequence being plant images of the target plant collected from a plurality of angles; inputting the multi-view image sequence into a target detection model to obtain a plant positioning image set and a plant positioning frame output by the target detection model; taking the plant positioning frame as prompt input, and inputting the plant positioning image set into a preset image segmentation model to obtain a plant mask of the target plant output by the image segmentation model; performing mask processing on the multi-view image sequence based on a plant mask to obtain a pure multi-view image; and performing three-dimensional reconstruction based on the pure multi-view image through a 3D Gaussian scattering framework to obtain an initial three-dimensional plant model corresponding to the target plant. According to the invention, the interpretability and the three-dimensional contour definition of the plant structure can be improved.

A comprehensive spatial AI platform for neural 3D reconstruction and built environment analysis integrates foundation models, deep learning methods, and spatial reasoning capabilities to provide expert knowledge of the physical world. The platform combines symbolic AI and machine learning to facilitate 3D semantics for insurance, real estate, construction, robotics, and other business applications. The system processes captured images, videos, or point clouds through neural networks with low compute requirements, incorporating device-agnostic advanced spatial intelligence that delivers geometric, semantic, and relational data. The platform includes proprietary training innovations, comprehensive measurement and semantic understanding, external sensor integration, human-in-the-loop quality assurance, and API integration for programmatic access. Advanced spatial reasoning capabilities enable property damage assessments, construction progress tracking, robotics navigation, and real-time applications including room dimension validation and automated repair estimates, supporting enterprise workflows across various industry segments while enabling productivity improvements and new value-added spatial AI use cases.

The invention discloses a binocular line laser reconstruction method and system, and relates to the technical field of computer vision, and the method comprises the following steps: S001, calibrating an internal reference and an external reference of a binocular camera, and obtaining a left image and a right image; arranging a line laser to respectively project parallel light knife surfaces on the left and right images; s002, respectively carrying out sub-pixel-level center extraction on the left image and the right image to obtain a binary mask, and calculating a candidate intersection point set; s003, calculating a basic matrix, determining a corresponding point on the right image based on the intersection point of the left image, and outputting a matching pair; s004, triangularizing the matching pair to obtain a 3D intersection point, and then marking the attributes of the light knife surface; s005, region growing is carried out after initialization; s006, calibrating a smooth knife surface equation; and S007, laser stripe full-pixel matching is carried out, and a final 3D point cloud is output. According to the invention, two or more groups of parallel line lasers in different directions are simultaneously utilized in the binocular system, so that unambiguous and high-density three-dimensional reconstruction is realized, the scanning efficiency is improved, and the ambiguity problem existing in the prior art is solved.

The invention relates to the technical field of moisture prediction, in particular to a 3D reconstruction and illumination adaptive coal surface moisture prediction method, which comprises the following steps of: planning an unmanned aerial vehicle route to collect an image, reconstructing coal pile surface space structure information, forming a local triangular grid, dividing an illumination continuous area, establishing a space corresponding relation between the illumination area and a hyperspectral pixel, and predicting the moisture content of a coal pile. And geometric registration and reflection transition correction are executed, depression points and shoulder positions of a reflection curve are detected in the identification interval, moisture absorption characteristics are constructed, a moisture migration link is established in the slope direction, and a coal surface moisture prediction result is obtained. According to the method, a spatial structure and illumination distribution are coupled, a dynamic correction relation is established to unify reflection information, image registration and spectrum reconstruction are combined to strengthen geometric consistency, a moisture migration trend is extracted by utilizing depression point and slope correlation, link and accumulation distribution is formed, and a corrected reflection characteristic represents a spectrum attenuation rule. The spatial continuity and numerical stability of moisture identification are ensured, and the water content grade division is coherent and comparable.

Optical sensors (e.g., cameras) and (e.g., IR) illumination sources may be distributed in an environment (e.g., an interior space such as a cabin or cockpit of an ego-machine) and synchronized to generate frames of sensor data, which may be used to reconstruct 3D geometry and / or 3D pose of an occupant, operator, or other object in the environment. For example, stereo vision may be used to generate one or more depth maps from image data generated using different cameras, the depth map(s) may be transformed into a 3D point cloud, and surface reconstruction may be applied to reconstruct the 3D geometry of surface(s) in the environment. A 3D pose, one or more keypoints (e.g., facial landmarks), or some other representation of the shape of the reconstructed surface(s) may be extracted from the reconstructed surface and used in one or more downstream tasks, such as driver and / or occupant monitoring tasks.

The present application provides a method and apparatus for three-dimensional (3D) reconstruction of a scene, an electronic device, and a storage medium. The method includes: obtaining a two-dimensional (2D) image and 3D point cloud data of a scene to be reconstructed, where the 2D image is a foreground image including a specified object; identifying an object in the 2D image, obtaining 2D data of the object, and performing 3D reconstruction based on the 2D data to obtain first feature data; performing point cloud segmentation on the 3D point cloud data to obtain object 3D point cloud data and background 3D point cloud data; performing 3D reconstruction based on the object 3D point cloud data to obtain second feature data, and performing 3D reconstruction based on the background 3D point cloud data to obtain third feature data; fusing the first feature data and the second feature data to obtain fusion feature data; and rendering the third feature data and the fusion feature data into a virtual space based on a position relationship between the object and a background to realize 3D reconstruction of the scene to be reconstructed. The method can improve the accuracy of 3D reconstruction of the scene.