Next: 1 Introduction Up: Contents

Visual Tracking of Solid Objects
Based on an Active Contour Model

Abstract:

A method for tracking the pose of solid objects moving with 6 degrees of freedom in space is proposed.

The novel idea is to split the tracking into two quasi autonomous working processes, one to track the 2D silhouette of the object in an image sequence and one to track the full 3D object pose based on the estimated 2D silhouette. At a lower level an active contour model utilizing a Kalman filter is used to efficiently track the object silhouette represented as contrast edges in the image sequence. Geometric information used are limited to a curve, where the current silhouette is assumed to lie within the affine space of this curve. At a higher level a 3D pose tracker derives the object pose from the estimated shape of the silhouette utilizing an geometric 3D object model. The pose is computed from an n-point correspondence between the silhouette and the object's surface in case of polyhedra and using a non-linear optimization method in case of smooth curved objects. Furthermore, the pose tracker predicts the object appearance, i.e. the object aspect, and provides the contour tracker with new model information on object aspect changes.

The presented results show the successful application of the method to smooth curved objects in synthetic image sequences.

• 1 Introduction
• 2 Active Contour
• 3 Solid Objects
• 4 Tracking the 3D object pose
• 5 Experimental Results and Concluding Remarks
• References