Wide-Area Egomotion Estimation from Known 3D Structure
Author(s)Koch, Olivier; Teller, Seth
Robotics, Vision & Sensor Networks
MetadataShow full item record
We describe an algorithm that takes as inputs a coarse3D model of an environment, and a video sequence acquiredwithin the environment, and produces as output an estimateof the cameraÂs 6-DOF egomotion expressed in the coordinatesof the 3D model. Our method has several novelaspects: it performs line-based structure-from-motion; italigns the local line constellation to the known model; andit uses off-line visibility analysis to dramatically acceleratethe alignment process.We present simulation results demonstrating themethodÂs operation in a multi-room environment. We showthat the method can estimate metric egomotion accuratelyand could be used for for many minutes of operation andthousands of video frames.
Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory
computer vision,visual geometry,omnivision,3D