Online Adaptive Decision Fusion Framework Based on Entropic Projections onto Convex Sets with Application to Wildfire Detection in Video

摘要

In this paper, an Entropy functional based online Adaptive Decision Fusion(EADF) framework is developed for image analysis and computer visionapplications. In this framework, it is assumed that the compound algorithmconsists of several sub-algorithms each of which yielding its own decision as areal number centered around zero, representing the confidence level of thatparticular sub-algorithm. Decision values are linearly combined with weightswhich are updated online according to an active fusion method based onperforming entropic projections onto convex sets describing sub-algorithms. Itis assumed that there is an oracle, who is usually a human operator, providingfeedback to the decision fusion method. A video based wildfire detection systemis developed to evaluate the performance of the algorithm in handling theproblems where data arrives sequentially. In this case, the oracle is thesecurity guard of the forest lookout tower verifying the decision of thecombined algorithm. Simulation results are presented. The EADF framework isalso tested with a standard dataset.