Visual Registration for Geographical Labeling in Wearable Computing

Real-time labeling of geographical landmarks is a simple wearable computing application. Most current systems are based on calculation of optical flow between the current and previous frames to adjust the label position. Here we present two alternative algorithms based on geometrical image constraints .The first is based on epipolar geometry and provides a general description of the constraints on image flow between two static scenes. The second is based on the calculation of a homography relationship between the current frame and a stored representation of the scene. A homography can exactly describe the image motion when the scene is planar, or when the camera movement is a pure rotation, and provides a good approximation when these conditions are nearly met. We assess all three styles of algorithm across a number of criteria including robustness, speed and accuracy. We also consider issues of representation and storage for geographical labeling applications.