Name
Abstract | ||
|---|---|---|
We are considering the problem of dynamic localization of human targets in an indoor environment, such as an office building, where GPS signals are not receivable. Previous work has shown that static localization is possible through the measurement of the wireless signal strengths. Dynamic localization (tracking) can be achieved by performing periodic static localizations and filling in the gaps through an appropriate filtering technique. We are using a sampling-importance-resampling particle filter which is a probabilistic reasoning technique for this purpose. In this paper we present approaches through which information about the environment (such as the floor plan of the building) and the target (such as the physical and social limitations of the human movement) can be incorporated in the prediction and weight update components of the particle filter. The particle filter requires information to be presented as conditional probability distributions, a format which presents both representational and computational efficiency challenges. In addition, the models need to consider both the a priori information and the operational details of the particle filter. Even technically correct models can reduce the accuracy of the localization, by inadvertently reducing the effective number of particles, which, in its turn, induces resampling errors. Through a series of experiments we show that the correct usage of a priori information can significantly improve the accuracy of dynamic localization. |
Year | DOI | Venue |
|---|---|---|
2009 | 10.1145/1641804.1641872 | MSWiM |
Keywords | Field | DocType |
human target,periodic static localization,indoor dynamic localization,correct model,human movement,correct usage,indoor environment,sampling-importance-resampling particle filter,particle filter,dynamic localization,a-priori information,static localization,sampling,tracking,particle filters,localization,conditional probability,signal strength,probabilistic reasoning | Computer science,Floor plan,A priori and a posteriori,Particle filter,Artificial intelligence,Probabilistic logic,Distributed computing,Computer vision,Conditional probability,Algorithm,Filter (signal processing),Monte Carlo localization,GPS signals | Conference |
Citations | PageRank | References |
2 | 0.40 | 15 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Begümhan Turgut | 1 | 36 | 4.93 |
| Richard P. Martin | 2 | 1777 | 165.29 |