Feature Point Analysis System

Computer Science & Math, Inventions & Innovation
Christopher Nielsen – Canada

Optical systems for tracking mobile platforms in three dimensions are typically limited due to cost and computational complexity. In this project, the hypothesis was that an inexpensive webcam stereoscopic system could be developed …

to estimate and track the state variables describing the position and orientation of a mobile platform. This was achieved through real time processing of feature points on the ground surface, observed by a stereoscopic vision system collocated with the platform. The researcher implemented an algorithmic tracking loop to optimize each state variable simultaneously and minimize errors between feature points on the ground. The result is an effective, inexpensive alternative to current geo-positioning systems.

Google Science Fair