714 results about "Human-body model" patented technology

The invention belongs to the technical field of three-dimensional reconstruction, and discloses a three-dimensional digital human generation method and system capable of voice interaction. According to the invention, brand new speaking audios in different languages are automatically generated according to different languages of the input target text and the sampled human voice audios; the sequential stability and detail reduction capability of three-dimensional human motion are guaranteed by using multi-model joint estimation and a sequential loss function, and facial expression details and hand postures in the image can be accurately estimated. After the high-precision three-dimensional human body model is obtained through estimation, human body action and expression generation is carried out based on voice driving, accurate synchronization of actions and expressions generated through voice is achieved, and facial expression movement and body posture movement, namely a whole-body three-dimensional human body model, conforming to brand-new speaking audio are accurately generated; and finally, rendering the whole-body three-dimensional human body model into a real digital human capable of voice interaction by using a three-dimensional neural rendering model. According to the invention, the realization of single person picture input, high-precision three-dimensional digital person generation and voice interaction is facilitated.

The invention relates to a single-image dressed human body reconstruction method based on posture guide diffusion and view angle consistency constraint, and belongs to the field of computer vision and graphic images. The method comprises the following steps: firstly, carrying out multi-view image generation on a single image in an input dressed human body data set by adopting posture guide diffusion to obtain a two-dimensional diffusion image of an invisible view angle; secondly, establishing a visual angle consistency constraint module, taking the input image and the diffusion image as bidirectional visual angle features, performing implicit geometric guidance in the normal direction of the SMPLX human body template, and respectively extracting three-dimensional space features corresponding to cross visual angles; and finally, fusing voxelization features and three-dimensional space features in the SMPLX human body template to carry out human body implicit reconstruction to obtain a dressed human body model. According to the method, the problems of low reconstruction quality, clothing detail missing and the like caused by insufficient input information of a single image can be effectively solved, and high-quality three-dimensional reconstruction of a dressed human body is realized.

The invention relates to the technical field of computer graphics and artificial intelligence, and discloses a meta-universe digital human generation method and system based on deep learning, and the method comprises the steps: constructing a neural radiation field human body model through multi-view video data, and learning a user behavior sequence mode; collecting sparse sensor data of the head and the hands of the user and generating a preliminary whole-body skeleton posture; calculating contact force feedback when the digital human interacts with the virtual environment through a differentiable physical engine; utilizing a reinforcement learning algorithm to optimize postures to meet physical constraints; performing multi-objective optimization and quality evaluation on the generated attitude; according to the method, the digital human postures which have personalized features and can naturally interact with the environment can be generated by using the consumer-level equipment, and the meta-universe interaction experience is improved.

The invention discloses a garment generation model training method, a digital human garment generation method and hardware. The method comprises the following steps: acquiring a three-dimensional human body model of a three-dimensional digital human to be dressed; obtaining first section data, and processing the first section data based on the three-dimensional human body model to obtain a three-dimensional grid model of the initial garment and second section data corresponding to the three-dimensional grid model; acquiring clothing image data including clothing texture information, and generating target image data including target clothing texture information based on the clothing image data and the second section data; the target image data is used for generating three-dimensional clothes data for dressing the three-dimensional digital human with the three-dimensional grid model; and obtaining a description text which corresponds to the target image data and is used for describing the clothing, and training the clothing generation model based on the target image data and the corresponding description text to obtain a trained clothing generation model.

The invention provides a 3D modeling-assisted garment pattern making method and system, and relates to the field of garment design, and the method comprises the steps: building a fabric database according to the fabric data on the current market, and obtaining an initial 3D garment model of a garment; fitting simulation is conducted on the obtained initial 3D garment model on the human body model, parameters of the initial 3D garment model are adjusted according to the simulation result and the fabric data, and an optimal 3D garment model is obtained; splitting the optimal 3D garment model to obtain a corresponding cut-part model and a sewing sequence; cloth is cut according to the cut-part model, solid cut-parts are obtained, the solid cut-parts are sewn according to the sewing sequence, manufacturing and adjusting of the clothes are completed, the pattern is confirmed, and therefore pattern making of the clothes is completed. According to the 3D modeling assisted clothing pattern making method and system, high precision, high efficiency and low cost of clothing pattern making are achieved, and meanwhile the design freedom degree and the wearing comfort degree are improved.

The invention provides a three-dimensional digital human reconstruction method, and the method comprises the steps: estimating a parameterized human body model corresponding to an input single human body reference image based on the input single human body reference image, and obtaining a multi-view human body semantic segmentation image of the parameterized human body model according to a plurality of preset camera poses and rendering parameters; taking the multi-view human body semantic segmentation image as a condition signal, taking the single human body reference image as input, and generating a multi-view human body color image based on a pre-trained first video diffusion model; taking the multi-view human body color image as a condition signal, taking a human body normal image extracted from a single human body reference image as input, and generating a multi-view human body normal image based on a pre-trained second video diffusion model; and performing 3D Gaussian splashing based on the multi-view human body color image and the multi-view human body normal image so as to reconstruct a three-dimensional digital human body corresponding to the single human body reference image.

The invention discloses an intelligent guiding method based on a 3D virtual garment customization process, and relates to the technical field of garment customization, and the method comprises the steps: obtaining the basic information of a user through a user interaction interface, and generating a 3D virtual human body model matched with the body type of the user through 3D scanning equipment; performing intelligent matching through a machine learning algorithm according to the basic information of the user and the style information in the costume design database, and displaying a matching recommendation result on a user interface in a visual form for the user to select and preview; dynamically adapting the 3D model of the clothes selected by the user to the 3D virtual human body model, and providing an intelligent adjustment option and an intelligent algorithm to generate an adjustment suggestion; according to the method and the device, fitting simulation, comfort evaluation and visual effect evaluation feedback are performed, user design is optimized, and a customization scheme containing a 3D model file and a production drawing is generated according to a design scheme finally confirmed by a user, so that the operation of a virtual garment customization process is simpler, the auxiliary decision-making experience degree is higher, and the customization efficiency is improved.

The invention discloses a point cloud compression method based on multi-dimensional feature fusion, electronic equipment and a medium, and aims to effectively reduce the compression bit rate while maintaining the precision of a space structure. According to the method, space, channel and topology redundant information is fused under a unified framework for the first time, and the method is suitable for efficient point cloud compression of various three-dimensional scenes such as automatic driving and virtual reality. Meanwhile, a local graph convolution Mama module is introduced, and the limitation that a point cloud topological structure is difficult to capture in a traditional method is effectively solved. Then, the octree nodes are divided into two dimensions of space and channel through a space-channel coupling grouping module, so that gradual feature fusion is realized, and the context information prediction precision is improved; experimental results show that the compression efficiency is remarkably improved on a plurality of radar and voxel human body model data sets.

The invention provides a breast endoscopic surgery simulation training system. The breast endoscopic surgery simulation training system comprises a hardware end and a software end, the hardware end comprises a simulation human body model, and the software end is provided with a 3D mesh model corresponding to the simulation human body model; the simulation surgical instrument comprises a simulation electrotome, a simulation endoscope, simulation grasping forceps and a simulation incision protection sleeve, and the software end is provided with an electrotome model, an endoscope model, a grasping forceps model and an incision protection sleeve model which correspond to one another; the pose detection device is used for detecting the current pose of the simulation surgical instrument in real time and sending the current pose to the software end, so that interaction between the simulation surgical instrument and the simulation human body model is mapped to interaction between the instrument model and the 3D mesh model; the software end further comprises a display screen, and the display screen is used for displaying interaction between the instrument model and the 3D mesh model in real time. According to the scheme, breast endoscopic surgery training, teaching and research which are closer to a real surgery scene can be provided, so that the safety and the operation precision in an actual breast endoscopic surgery are improved.

The invention belongs to the technical field of medical training, and particularly relates to a method and a system for constructing a virtual operating room resource library for reappearing an offshore operation scene, and the method comprises the following steps: 1, collecting offshore medical data and scene demands: collecting an offshore environment and medical data, and determining medical care core capability demands; in the application, coupling parameters generated by a sea condition dynamic sensing module drive a mixed reality scene rendering module to present real shake and environment change; meanwhile, parameters are synchronously input to the equipment modeling module, fault response of virtual equipment is triggered, a user operates the virtual equipment through the interaction module and executes an operation on a virtual human body model, and the resource library management module guarantees reusability and updatability of resources in the whole process. Through real-time data linkage among the modules, total-factor dynamic coupling of sea conditions, scenes, equipment and operation objects is finally realized, and the problem of simulation distortion of a traditional system is solved.

The invention relates to a basketball player motion capture analysis method and system based on an industrial camera. The method comprises the following steps: based on a preset three-dimensional human body model, obtaining a three-dimensional human body surface grid sequence and internal skeleton joint kinematics data by combining a pixel-level mask of each multi-view image frame set; detecting a body contact event and a continuous time period based on the Euclidean distance of the three-dimensional human body surface grid sequence; based on the internal skeleton joint kinematics data, calculating a multi-dimensional kinematics feature time sequence; constructing a dynamic causal hypothesis based on the body contact event, checking the dynamic causal hypothesis based on a preset differentiable forward dynamic multi-rigid-body human body model and a multi-dimensional kinematics characteristic time sequence to obtain a loss function value, and judging the hypothesis with the smaller loss function value as a real dynamic causal relationship. And the identity identification information is combined to obtain the identity of the active force generation party and the identity of the passive force bearing party. By adopting the method, a real force exerting main body and a real force bearing main body in confrontation can be accurately identified.

The invention relates to the cross technical field of medical education and mixed reality technology, in particular to a traditional Chinese medicine meridian point visualization system based on mixed reality technology, which comprises a data fusion and personalized calibration module, a virtual-real fusion interaction module, an interaction teaching module, an intelligent evaluation module and a meridian qi and blood simulation module. By fusing medical image data, depth sensing data and a traditional Chinese medicine acupoint selection rule, a personalized three-dimensional human body model is constructed, and acupoints are accurately calibrated; real-time registration and shielding processing of the virtual model and a real scene are achieved through mixed reality equipment; various natural interaction modes such as gestures, voices and sight lines are supported, and an integrated function of learning, training and assessment is provided; and the meridian qi-blood running state can be dynamically simulated based on the midnight-noon ebb-flow theory. The system has the advantages that the problems that traditional Chinese medicine teaching is poor in intuition and interactivity, cannot be personalized, is inconvenient to examine and the like are solved, and immersive, interactive and intelligent teaching of traditional Chinese medicine meridian point knowledge is achieved.

The invention discloses an intelligent human body modeling method and system based on a three-dimensional dot matrix, and relates to the field of intelligent modeling, and the method comprises the steps: carrying out the structural analysis of the semantic level of an original three-dimensional point cloud, and predicting a space displacement field pointing from the surface of observed clothes to the surface of an internal real human body under the semantic guidance. By applying the displacement field, double transformation of point cloud stripping and standardization is carried out on the point cloud, so that repeated fine tuning and optimization are carried out on the body type and posture parameters of the human body model in a standardized space, and finally a high-fidelity human body three-dimensional model is generated through driving. In this way, the complex clothes interference problem is deconstructed into a staged geometric mapping and parameter optimization process, and therefore the robustness and accuracy of human body three-dimensional reconstruction under the real wearing condition are effectively improved.

The invention discloses a virtual image clothes changing method and system, and the method comprises the steps: carrying out the preprocessing of a figure image inputted by a user, and obtaining standardized image data; recognizing a face region in the standardized image data, and constructing a three-dimensional face model; body parameters corresponding to the three-dimensional face model are determined, and a three-dimensional human body model is generated based on the body parameters; based on a matching result between the measured size of each key part in the three-dimensional human body model and a design size of a corresponding part in a preset clothing template, deforming the clothing template to obtain a three-dimensional clothing model, and simulating virtual clothing changing effect data when the three-dimensional human body model wears the three-dimensional clothing model; and updating the three-dimensional human body model according to the facial features of the user and the body posture of the user, updating the virtual clothes changing effect data according to the updated three-dimensional human body model, and obtaining a virtual image preview picture corresponding to the dynamic performance behavior of the user.

The invention provides a single-image reconstruction method of an animatable digital human based on three-dimensional Gaussian splashing, and the method comprises the steps: inputting a single human body image into a preset three-dimensional perception diffusion model and a preset SMPL-X parameterized human body model, and determining a multi-view image and a standard human body grid; a preset three-plane encoder and a preset multi-layer perceptron are adopted to predict three-dimensional Gaussian primitive attributes and point-to-point correspondence between three-dimensional Gaussian primitives and vertexes of the standard human body grid for the standard human body grid; adopting a preset target parameter to drive and render the three-dimensional Gaussian primitive, and determining a digital human image under the target parameter; determining an optimized three-plane encoder and an optimized multi-layer perceptron; and determining a target digital human image by adopting the optimized three-plane encoder and the optimized multi-layer perceptron. According to the method and the device, the high-quality three-dimensional digital human is reconstructed by adopting a single image, the data dependence is reduced, the data acquisition process is simplified, and the three-dimensional digital human making efficiency is improved.

The invention relates to the cross technical field of robot control, computer graphics and motion capture processing, in particular to a humanoid robot motion migration method and system based on motion capture, and the method comprises the following steps: S1, collecting human motion data and converting the human motion data into a general format, specifically, recording the motion trail of a human mark point through a motion capture system, the method comprises the following steps: generating a. Bvh file containing skeleton hierarchy and motion parameters, converting the. Bvh file into an NPZ format file containing translation and posture parameters, and processing the NPZ format file based on an SMPL-X human body model. According to the method, through scale optimization and joint matching, the problem of scale difference between the SMPL human body model and the robot model is solved, the matching precision of the joint positions is improved, through combination of the optimization strategy of the main loss and the regularization item and physical constraints, the robot joint angle is effectively prevented from exceeding the limit, and the physical feasibility of actions is ensured.

The present invention discloses a method and an apparatus for rebuilding a relightable implicit human body model. A human body is represented as a deformable implicit neural representation, and a geometric shape, material attributes, and ambient lighting of the human body are decomposed to obtain a relightable and drivable implicit human body model. In addition, a volumetric lighting grid including a plurality of spherical Gaussian models is introduced to represent complex lighting with spatial variation, and visible probes capable of changing in position with change in human pose are introduced to record dynamic self-occlusion caused by human motion. With the method, drivable implicit models capable of being used for high-fidelity human body relighting may be generated in cases of sparse video input and even monocular input.

The invention relates to the technical field of protective clothing testing, and discloses a protective clothing combustion testing device based on an intelligent sensor, which comprises a control console, a testing cabinet is arranged on the control console, a simulation mechanism for simulating real motion is arranged on the inner wall of the testing cabinet, and an exhaust mechanism for ventilation is arranged on the inner wall of the testing cabinet. According to the device, in the rotating process of the fire protection clothing, it can be ensured that heat flow of subsequent flame testing is uniform, testing and detection are free of dead corners, the real wearing state is simulated, the flame testing quality of the device on the fire protection clothing is improved, the testing combustion precision of the device on the fire protection clothing is improved, and the testing efficiency is improved. And meanwhile, a temperature sensor on the surface of the human body model can test the internal temperature of the fire protection clothing, all-directional combustion test detection of the device on the fire protection clothing can be improved through rotary movement of a flame gun, and the combustion test effect of the device is improved.

The invention relates to the technical field of computer vision, e-commerce and artificial intelligence, and discloses a computer vision-based online clothing size intelligent matching and fitting degree prediction method, which comprises the following steps of: firstly, reconstructing a high-precision three-dimensional human body model by processing an image uploaded by a user; meanwhile, three-dimensional sizes and physical attributes are deconstructed from a clothing commodity graph, then, a physical simulation engine is utilized to dynamically wear a clothing model on a user model for virtual fitting, the wearing effect is truly restored by simulating physical effects such as gravity and collision, quantitative fitting indexes such as tightness and oppression are calculated after simulation, and the fitting quality is improved. And based on multi-objective optimization, scoring different sizes, recommending an optimal size, and generating a visual report containing the 3D effect picture and the fitness thermodynamic diagram. According to the invention, automation and precision from the image to fitting degree prediction are realized, the problem of online clothes purchase size selection is effectively solved, the user experience is significantly improved, and the return rate is reduced.

The invention relates to the technical field of computer vision, and particularly provides a virtual digital human generation method and electronic equipment, and the method can comprise the steps: obtaining three-dimensional human body modeling data corresponding to each frame of image in a monocular video sequence of a target object; the three-dimensional human body modeling data comprises body posture data and identity offset data; processing the three-dimensional human body modeling data by using a pre-constructed to-be-trained virtual model to generate a virtual object image of each frame of image; the to-be-trained virtual model comprises a to-be-optimized three-plane feature body and a plurality of multi-layer neural network MLP models; optimizing the to-be-trained virtual model through the loss value between the virtual object image and each corresponding frame image, and obtaining a three-dimensional digital human model corresponding to the target object; the three-dimensional digital human model is used to generate a virtual digital human based on any drive data. According to the embodiment of the invention, the generation and rendering quality of the virtual digital human can be improved.

The invention discloses an intelligent clothing design and template generation method and system, and relates to the technical field of clothing intelligent plate making, and relates to multi-physics field modeling, dynamic simulation, reverse optimization and iterative output technologys.The method comprises the steps that a biomechanical human body model containing skeletal muscle groups and a multi-physics field coupling model of initial clothing piece templates and fabric parameters are constructed; defining a target pressure curve; driving a human body to execute preset actions, and calculating a dynamic pressure field and a wrinkle discomfort index through explicit dynamics; constructing a loss function fusing the pressure goodness of fit and the wrinkle index, solving a gradient by using an adjoint method, and accelerating reverse optimization in combination with a physical information neural network agent model; the system comprises corresponding function modules, parameters can be automatically adjusted according to the motion posture, it is ensured that the dynamic pressure is always close to a target curve, and the functionality and comfort level of clothes are improved.

The application provides a virtual reality display method and system of a rehabilitation robot based on image segmentation, and relates to the technical field of virtual display. The method can determine an intermediate 3D human body model according to an external body feature vector corresponding to a target user and an external body feature vector corresponding to a preset 3D human body model, determine a target 3D human body model according to a first bone feature vector corresponding to the target user and a second bone feature vector corresponding to the intermediate 3D model, drive the target 3D human body model to perform synchronous movement according to decomposed actions completed by the target user when performing a training task, and mark each decomposed action using a difficulty level label. The method effectively reduces the influence of external factors such as clothing and accessories of the target user on data collection and processing, can more accurately restore the real action details of the target user, is beneficial to improving the user experience, avoids a complex reconstruction process of three-dimensional point cloud data, and reduces the consumption of computing resources.

This invention relates to the fields of computer vision and computer graphics, and provides a method and system for dynamic human body reconstruction from monocular video based on 3D Gaussian splashing. The method includes the following steps: data preprocessing; initialization of the normalized space 3D Gaussian; deformation of the normalized space 3D Gaussian to an intermediate pose space to obtain a non-rigidly deformable 3D Gaussian and pose-related features; transformation of the non-rigidly deformable 3D Gaussian to the observation space using linear blending skinning to obtain the observation space 3D Gaussian; decoding the viewpoint-related color based on Gaussian color features, pose-related features, and viewpoint direction; constructing a total loss function including a normal consistency regularization term to optimize the 3D Gaussian attributes and network parameters; and rendering the target human body image using a differentiable Gaussian splash rasterizer. This invention enables rapid and fully automatic reconstruction from monocular video to a high-fidelity, animable human body model, applicable to fields such as virtual reality and film production.

1. Name of the product in this design: Bra (2609). 2. Purpose of this design: Women wearing bras. 3. The key design features of this product are the combination of shape and pattern. 4. The image or photograph that best illustrates the design's key points: a 3D model. 5. The design points for the top view have been shown in other views, so the top view is omitted; the design points for the bottom view have been shown in other views, so the bottom view is omitted. 6. Other situations requiring explanation: The human body model is used to support the clothing and is not within the scope of protection of this design.

1. Name of the product in this design: Underwear (F250807). 2. Purpose of this design: for everyday wear. 3. The key design feature of this product is its shape. 4. The image or photograph that best illustrates the design's key points: a 3D model. 5. The key design points of the product from the top view have been shown in other views, so the top view is omitted; the bottom surface of the product in this design is a part that is not easily seen or cannot be seen during use, so the bottom view is omitted. 6. Other situations requiring explanation: The human body model is used to support the clothing and is not within the scope of protection of this design.

1. Name of the product in this design: Bra (KW8680). 2. Purpose of this design: For women to wear bras. 3. The key design features of this product are the combination of shape and pattern. 4. The image or photograph that best illustrates the design's key points: a 3D model. 5. The design considerations for the top view of the product have been shown in other views, so the top view is omitted; the bottom view of the product is generally not easy to see during use, so the bottom view is omitted. 6. Other situations requiring explanation: The human body model is used to support the clothing and is not within the scope of protection of this design.

The application relates to the technical field of medical data processing, and discloses a macroscopic cross-section database construction method, a dose determination method, a device and equipment, which comprises the following steps: setting multiple boron concentration values for the tissue types corresponding to a human body model; selecting corresponding neutron source parameters for the tissue types and the multiple boron concentration values, wherein the neutron source parameters comprise energy range values; dividing the energy range values into multiple discrete neutron energy groups; performing particle transport simulation based on the selected neutron source parameters, jumping between different neutron energy groups in the simulation process, and obtaining neutron reaction rates and neutron fluxes of the tissue types under the multiple boron concentration values and the multiple neutron energy groups; and obtaining neutron macroscopic cross-section values of the tissue types under the multiple boron concentration values and the multiple neutron energy groups based on the neutron reaction rates and the neutron fluxes, and constructing a multi-group macroscopic cross-section database. The application can improve the calculation efficiency of the radiation dose and improve the adaptability to individual metabolic differences during dose calculation.

1. Name of the product in this design: Bra (6628-6629). 2. Purpose of this design: For women to wear bras. 3. The key design features of this product are the combination of shape and pattern. 4. The image or photograph that best illustrates the key design points: Design 1 3D view. 5. The design points for the top view of Design 1 are shown in other views, so the top view of Design 1 is omitted; the design points for the bottom view of Design 1 are generally not easy to see during use, so the bottom view of Design 1 is omitted; the design points for the top view of Design 2 are shown in other views, so the top view of Design 2 is omitted; the design points for the bottom view of Design 2 are generally not easy to see during use, so the bottom view of Design 2 is omitted. 6. Design 1 is designated as the basic design. 7. Other situations requiring explanation: The human body model is used to support the clothing and is not within the scope of protection of this design.

The invention discloses a dynamic wearing restoration method and system for ancient clothes. The method comprises the following steps: collecting ancient garment data to obtain a cut-part picture; establishing a human body model according to the human body data of the garment wearer; based on the obtained cut-part picture, establishing a clothing model; the method comprises the following steps: acquiring motion capture data through motion capture equipment, importing the motion capture data into Marveless Designer software, and carrying out cloth dynamic calculation in combination with a human body model and a clothing model to complete dynamic simulation; using Adobe Photoshop software to make a texture map of the clothing, and completing visual optimization of the clothing; dynamic display is achieved through Maxon Cinema 4D software, and high-precision rendering is conducted through a Redshift renderer; the restoration of the dynamic wearing process of the ancient clothes is completed. The three-dimensional dynamic restoration of the ancient clothes is realized, and the dynamic wearing effect of the ancient clothes is truly restored.

The invention provides an intelligent binding skill operation training robot and a control method. The operation training robot comprises a simulation human body model, a dynamic driving system, a wound simulation module, a sensing matrix system and a control system. The simulation human body model, the dynamic driving system, the wound simulation module, the sensing matrix system, the control system and the artificial intelligence evaluation system are combined to construct an intelligent simulation training platform which is highly vivid, accurate in feedback and capable of meeting various training requirements; according to the operation training robot, a passive static model is changed into an active interactive intelligent robot, the operation training robot is no longer only an operation object, but a simulated wounded person capable of sensing operation and dynamically responding according to an internal physiological model, and therefore a training environment with the sense of reality and challenging is provided for trainees, and the training efficiency is improved. And the operation skill, clinical decision and strain capability can be comprehensively exercised.