Research Projects

View Planning for Automated Site Modeling

We present a systematic method for constructing 3-D models of large indoor and outdoor sites. The method is designed for a mobile robot platform and incorporates automated acquisition of scanned data as well as automated view planning and model construction. In our modeling process, we first use a preliminary view or set of preplanned views to yield an initial, approximate, 3-D model of the target structure. Then, we update this model by using a voxel-based procedure to plan and acquire the next best view. This updating is repeated sequentially until an accurate and complete 3-D model is finally obtained. The method has been successfully tested on a portion of the Columbia University campus, Fort Jay at Governors Island, and several churches in the Bourbonnais region of France. Modeling Fort Jay at Governors Island, New York Data Acquisition and View Planning for 3-D Modeling Tasks Two Stage View Planning for Large-Scale Site Modeling View Planning for Automated Site Modeling Thesis Proposal - Automated Construction of Environment Models by a Mobile Robot Thesis Proposal Slides

Candidacy Exam - View Planning

For my PhD candidacy exam at Columbia University I presented a survey of the state of the art in view planning literature. Candidacy Exam Slides Candidacy Bibliography

Topological Localization for Mobile Robots

We developed a system for topologically localizing mobile robots. The first part of the system performs localization by matching multiresolution histograms of omnidirectional images to a database. The second reinforces the location estimate by matching measured wireless signal strengths measurements with a signal map of a known wireless network. A Hybrid Approach to Topological Mobile Robot Localization TopBot TopBot: Automated Network Topology Detection with a Mobile Robot Topological Mobile Robot Localization Using Fast Vision Techniques

3D Desktop Photography

We implemented 3D Photography on Your Desk, originally done by Jean-Yves Bouguet and Pietro Perona at Caltech. The goal of this system is to provide a simple and inexpensive way to extract the 3D shape of objects, using weak structured lighting. The only hardware necessary here is a video camera (with host computer), a checkerboard, a desk lamp and a pencil. Using the shape of the shadow cast by the pencil, the system should extract 3D points corresponding to the surface of the object over which the shadow passes. 3D Desktop Photography by Eclipse

Robot Path Planning Using Generalized Voronoi Diagrams

I implemented an efficient and robust algorithm for computing safe paths for a mobile robot. I have specifically designed the algorithm for use in the Autonomous Vehicle for Exploration and Navigation in Urban Environments (AVENUE) project. I use an approach based on the generalized Voronoi diagram for a planar region with specified obstacles. Once this diagram has been constructed, I can search it to find robot paths that pass, with maximal clearance, around the obstacles. Robot Path Planner A detailed report on the generalized voronoi diagram method that was used for the path planner. An applet demonstrating the system.

Autonomous Vehicle for Exploration and Navigation in Urban Environments (AVENUE)

AVENUE stands for Autonomous Vehicle for Exploration and Navigation in Urban Environments. The goal of this project is to create an autonomous system capable of building photo-realistic 3D geometrically accurate models of outdoor sites. In my early career at Columbia I worked on various low level parts of this project including GPS and camera integration. Many of my later projects all tie back into this project including the path planner and my thesis topic, the view planner. AVENUE: Automated Site Modeling in Urban Environments Design, Architecture and Control of a Mobile Site-Modeling Robot