Hierarchical and conditional combination of belief functions induced by visual tracking

In visual tracking, sources of information are often disrupted and deliver imprecise or unreliable data leading to major data fusion issues. In the Dempster-Shafer framework, such issues can be addressed by attempting to design robust combination rules. Instead of introducing another rule, we propose to use existing ones as part of a hierarchical and conditional combination scheme. The sources are represented by mass functions which are analysed and labelled regarding unreliability and imprecision. This conditional step divides the problem into specific sub-problems. In each of these sub-problems, the number of constraints is reduced and an appropriate rule is selected and applied. Two functions are thus obtained and analysed, allowing another rule to be chosen for a second (and final) fusion level. This approach provides a fast and robust way to combine disrupted sources using contextual information brought by a particle filter. Our experiments demonstrate its efficiency on several visual tracking situations.