Dense Reconstruction Using Fixation and Stereo Cues

摘要

In this paper, we investigate the issue of accurate estimation of the three-dimensional (3D) coordainates of a static scene from real images, combining fixation and stereo cues. We may need to compute 3D data in many applicaitons: vehicle positioning and maneuver, object observation and recognition, moving or fixed obstacle avoidance, 3D mapping for surveillance, etc. More specifically, we discuss the idea of using fixation to recover the 3D coordinates of some points in the robotic frame to help an uncalibrated camera to reconstruct a static scene.