Token tracking in a cluttered scene

摘要

The statistical data association technique is an important approach to analyse long sequences of images in computer vision. Although it has extensively been studied in other domains, such as radar imagery, it was introduced only recently in computer vision, and is already recognized as an efficent approach to solving correspondence and motion problems. This paper has two purposes. The first is to present a general formulation of token tracking. The parameterization of tokens is not addressed. This might be useful to those who are not familiar with statistical tracking techniques. The second is to introduce some strategies for tracking with emphasis on practical importance. They include beam search for resolving multiple matches, support of existence for discarding false matches, and locking on reliable tokens and maximizing local rigidity for handling combinatorial explosion. We have implemented those strategies in a 3D line segment tracking algorithm and found them very useful.