Onboard cameras capture facial features to map points onto 3D models, eliminating external markers for accurate avatar animation.
A system maps 3D object poses to associate image portions with surfaces and generate composite views.
A 3D scene generation method updates source image features using user editing instructions to enable interactive modifications.
A 3D imaging module combines mesh data with compensating information to maintain smooth partial-feature changes in virtual characters.
A system segments image streams into sub-groups to extract features and generate point correspondences for 3D reconstruction.
A machine learning model estimates a surface normal tensor from segmented image portions to apply augmented reality elements with pixel-level precision.
A dynamic angle viewing system generates real-time 3D models from depth and color sensors for multi-angle rendering.
Neural network segmentation of RGB camera images constructs a 3D user model via geometric transformation, eliminating dedicated motion sensing hardware.
Spatial object management engine generates integrated 3-D models from hand gestures to identify target objects.
Camera-based image recognition identifies grid patterns to construct 3D space grids, eliminating interference and crosstalk from active radar or laser sensors.
Automated feature matching replaces manual selection to reduce registration time while maintaining measurement precision in underground mapping.
A toy scanner captures radiation from multiple viewpoints to create accurate 3D digital representations of physical models.
Synthetic 3D model training enables a neural network to resolve sensor complexity constraints while maintaining real-time volumetric video fidelity.
A vision system tracks plant parts across multiple images using composite similarity scores.
Pre-computed voxel data structures assign optimal camera pairs to specific regions, resolving processing bottlenecks in free-viewpoint 3D reconstruction.
A dental restoration design system filters prosthetic variables using determined 3D surface anatomy.
Agricultural modeling system fuses 3D point cloud data with geospatial layers to generate multi-layered predictive models.
Segmented voxel processing overcomes view variance and occlusions in crowded environments.
A 3D color mesh reconstruction method corrects vertex color values to eliminate texture seams.
Ranking engine prioritizes target dimensions from 3D models to reduce supplementary measurements.
A system generates sharp 3D models by combining in-focus pixels from multiple 2D images captured with different depths-of-field.
A computer-assisted surgical system creates three-dimensional anatomical models using bi-planar two-dimensional images.
Hierarchical octrees with adaptive leaf resolution and parallel processing enable real-time reconstruction of large volumes while reducing memory usage.
Segmented simulation maps high-resolution textures onto simplified 3D garments, resolving the trade-off between rendering quality and computational time.
A statistical point pattern matching technique generates radial functions from aerial views to identify corresponding roof points.
Parallel cloud processing of preprocessed images reduces reconstruction time by 79% for large-scale agricultural project monitoring.
Private shadow graphics translation tables allow multiple vGPUs to share overlapped memory, overcoming scalability limits of three instances per physical GPU.
Segmenting depth images by linking information builds accurate 3D models of deformable objects with a single camera, resolving structural misalignment.
A 3D deformation model estimates pixel positions to perform non-rigid alignment between respiratory phase images.
Segmenting facial regions into local features improves detail restoration quality while managing computational complexity.
Local error bars guide adaptive smoothing of 3D laser scanner data, reducing statistical noise and false scatter points while preserving object shape accuracy.
An inflatable bladder fixture enables accurate 3D clothing measurement, resolving online fit uncertainty and reducing return rates.