Current Research Lines in 3D Computer Vision

• Acquisition

• Application

• Visualization

• Processing

Acquisition

• Vehicle-borne 3D scanning system

Back to 3D Computer Vision

Go to top

Application 

• Modular approach: acquisition, import file formats, pre-processing, 3D viewer, image calibration, scan registration…

• Pervasive out of core techniques to handle virtually any size of scans

• Octree space subdivision (including out of core) for efficient processing and visualization

• XML application configuration

Back to 3D Computer Vision

Go to top

Visualization

• Frustum culling and prefetching by means of octrees for point clouds and triangles

• Pixel Shaders 2.0 and Vertex Shaders 2.0: allows real time re-projection, occlusion query, some image processing and other special effects

Back to 3D Computer Vision

Go to top

Processing

• Effective registration and fusion of multi-sensory, real-world geographically referenced data (3D ground-level range data, video, GPS (et al.), 2D and 3D aerial images, satellite images, photos, thermal, X-ray images, etc.)

• Range scan conversion to Vector fields for efficient integration, hole filling, triangulation, registration, inspection

• Reconstruction of trajectory from GPS, INS points, and 2D profiles

Back to 3D Computer Vision

Go to top