Photogrammetry and 3-D Scene Reconstruction for Interactive Model-Based Video

Research by Shawn Becker, Ph.D. MIT Media Laboratory, 1997

A semi-automated process
that uses multiple photos of a scene ...


... to quickly create a textured 3-D model ...

... which can be used for interactive model-based video

 

About Shawn Becker

I'm an alum of the MIT Media Laboratory

I did my doctoral research under the advisement of professors
V. Michael Bove Ph.D.
Alex Pentland Ph.D.
Seth Teller Ph.D.

Print or read my Ph.D. dissertation online.

key words: computer vision, image processing, line detection, center of projection, image center, focal length, 3-D sphere of plane normals, orthogonality, covariance-based weighting, clutter avoidance, planar clustering, multi-sample fusion, parallel lines, vanishing points, distortion modeling, simplex cost minimization, interior camera modeling, photogrammetry, scene-relative camera orientation & position, error sensitivity, circle of confusion, inverse texture mapping, computer graphics, texture mapping, feature detection, feature tracking, camera path interpolation, model-based video coding, video compression, MPEG-4 multi-media layers, multi-layered audio and video, 3-D scene model, 3-D camera position, 3-D camera orientation, camera zoom, 3-D audio sources, lighting, actors, objects, 2-way internet communications, interactive video applications, interactive virtual camera and scene manipulation, links to actor and product information, e-commerce applications.


I'm currently co-founder and CTO of ClipFile.com

| home | overview | research | sbecker (at) alum (dot) mit (dot) edu |