2018/01/02 by Jérémie Houssineau, Houssineau, Jeremie
Computer Science · Environmental Science · #Data Management and Algorithms #FOS: Computer and information sciences #Methodology (stat.ME) #Remote Sensing in Agriculture #Soil Geostatistics and Mapping
paper · pdf · doi:10.48550/arxiv.1801.00571
openalex publication_date 2018/01/02 · openalex created_date 2022/08/22 · openalex updated_date 2026/07/28
We present a modelling framework for multi-target tracking based on\npossibility theory and illustrate its ability to account for the general lack\nof knowledge that the target-tracking practitioner must deal with when working\nwith real data. We also introduce and study variants of the notions of point\nprocess and intensity function, which lead to the derivation of an analogue of\nthe probability hypothesis density (PHD) filter. The gains provided by the\nconsidered modelling framework in terms of flexibility lead to the loss of some\nof the abilities that the PHD filter possesses; in particular the estimation of\nthe number of targets by integration of the intensity function. Yet, the\nproposed recursion displays a number of advantages such as facilitating the\nintroduction of observation-driven birth schemes and the modelling the absence\nof information on the initial number of targets in the scene. The performance\nof the proposed approach is demonstrated on simulated data.\n