The probability-hypothesis-density simultaneous localization and mapping filter is a random-finite-set estimation method that incorporates the probability-hypothesis-density filter within a Rao-Blackwellized particle filter, and was developed for navigation and mapping problems. However, the filter tends to diverge due to the existing importance-weighting methods used in the Rao-Blackwellized particle filter. This article introduces a new importance-weighting method that drastically improves the robustness of the probability-hypothesis-density simultaneous localization and mapping filter. Performance evaluations are conducted using both simulations and real experimental data sets.
Multifeature-based importance weighting for the PHD SLAM filter
IEEE Transactions on Aerospace and Electronic Systems ; 52 , 6 ; 2697-2714
01.12.2016
4804152 byte
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Corner detection algorithm based on multifeature [4550-38]
British Library Conference Proceedings | 2001
|Multifeature fusion using pulse-coupled neural networks
British Library Conference Proceedings | 1999
|