Details Involving Processing Steps TRIZ Cases
- Vascular Tree Reconstruction from Single 2D Projection
Surface model constraints recover 3D depth from a single 2D projection, eliminating noise from heartbeat and respiratory movement.
- Dynamic Facial Attribute-Based Special Effect Fusion
A system fuses target special effects by detecting facial attribute information of a target subject in video frames.
- Visual Localization Map Generation Using 3D Model Data
A visual localization system generates feature point maps by rendering images from virtual camera poses using 3D model data.
- Depth Map Generation Using PatchMatch and Poisson Refinement
PatchMatch generates initial maps while Poisson equations refine values to resolve texture-less region accuracy.
- Virtual 3D Model Creation via 2D Scan Vector Conversion
Transforms layered 2D scans into editable voxel models, resolving resolution limits in custom product fabrication.
- 3D Environment Reconstruction Using Planar Constraints
A mapping engine collapses 3D points into cohesive planar surfaces to reduce computational complexity.