Augmented Reality

Introduction

The Augmented Reality for Supervisory Rover Control task has developed technology that facilitates the control of rovers operating in distant environments. The approach augments operator displays by overlaying raw rover-acquired imagery with additional information such as topographic landmarks, heading and position estimates, and planned routes. We overlay this information in such a manner that it appears to "stick" to the terrain, independent of the rover's motion.

The task was succesfully completed during the course of the last quarter. The fourth quarter was spent in developments of image registration algorithms for image stabilization and in tests with data acquired previously.

We had finished the position estimation algorithm with excellent results in the third quarter. All results are available on the World Wide Web. Please refer to that website for more extensive document about the final status of the project.

A. Significant Activities and Events During this Period

The main effort during the last quarter was directed to:

the development and test of an image stabilization algorithm that can reduce the impact of image fast variations;

the consolidation of all system modules in a unified, final package.

The image stabilization module obtains the displacement between consecutive images through image registration techniques (phase correlation and iterative optimization) and filters the displacement, so that image motion is "smooth". The algorithm also allow us to overlay information in the incoming video stream, in case the user wants to verify the position of map and image features. The overlays can be "fixed" to points in the image, and the image registration algorithm guarantees that the overlay and the feature are associated as depicted in the image below.

This image may take a long time to download (approx. 1MByte)

B. Plans for the Next Reporting Period

None; task has been completed.

C. Schedule

D. Concerns/Issues

None.

Prepared by fgcozman@cs.cmu.edu