Vision and Mobile Robotics Laboratory | Videos

Home | Members | Projects | Publications | Software | Videos | Job opportunities

Internal

Videos

Automatic Photo Pop-up - automatically constructs simple "pop-up" 3D models, like those one would find in a children's book, out of a single outdoor image. [Details]

Volumetric Features for Recognition in Video - Event detection results using volumetric features. [Details]

Sample past projects

October 2004: Classification and Recognition - Object classification and recognition in images, video and 3-D data. [ Details]

Terrain classification

Wire detection - Linear structures are detected using local spatial point distribution [Details]

CTAC'04 video - Progress in terrain classification using 3-D ladar data [Details]

Recognition

Recognition using harmonic maps - Objects are recognized and 3D views are registered using a technique based on harmonic maps [Details]

Recognition using spin images - Objects are recognized and localized in cluttered 3D scenes using local surface signatures called spin-images. [Details]

Registration

View registration - Unconstrained alignment of pairs of 3D views of an object. [Details]

Modeling

Interior modeling - View registration can be extended to multiple views and used to model building interiors [Details]	Terrain modeling - A mobile range sensor is used to create a 3D map of an unstructured environment. [Details]
Automatic modeling - Given a set of views of an object or scene, we automatically reconstruct the model without any manual intervention. [Details]

---

The VMR Lab is part of the Vision and Autonomous Systems Center within the Robotics Institute in the School of Computer Science, Carnegie Mellon University.