Electronic Stabilization and Feature Tracking in Long Image Sequences

Abstract : This dissertation is concerned with processing of visual motion with application to off road vehicular navigation. We first consider the estimation of total rotation from an image sequence. We exploit the dynamic nature of the sequence and use multiple visual cues to perform image stabilization. Both calibrated and uncalibrated stabilization schemes are designed. The residual motion in a stabilized sequence is also analyzed. Next we address the issue of selective stabilization, defined as the separation of the smooth rotation and the residual vibrations caused by oscillatory rotation. We incorporate a kinetic model to explicitly account for vibration phenomena. A maneuver detection scheme, for detecting the beginning and end of smooth rotation, is designed to facilitate the selective stabilization. Finally, we study the problem of feature correspondence. We propose a localized feature point tracking algorithm, which employs a 2D kinematic model and relies on a Probabilistic Data Association Filter for the estimation of interframe motion. Corresponding points are identified to subpixel accuracy and an Extended Kalman Filter is employed to process the new data. The ability to dynamically include new feature points from subsequent frames also makes the algorithm suitable for structure from motion and tracking over a sequence.