Search
Abstract
The paper addresses some challenges pertaining to the methods for tracking of objects in multi-camera systems. The tracking methods related to a single Field of Vision (FOV) are quite different from inter-camera tracking, especially in case of non-overlapping FOVs. In this case, the processing is directed to determine the probability of a particular object’s identity seen in a pair of cameras in the presence of places non-observed by any camera, thus an object can disappear in one observed region and then re-appear in another one. A methodology for evaluation of the introduced re-identification method is presented in the paper. Problems related to the preparation of the ground-truth database and to the impact of a single-camera tracking on the efficiency of the re-identification algorithm are discussed.
Citations
-
0
CrossRef
-
0
Web of Science
-
0
Scopus
Authors (2)
Cite as
Full text
full text is not available in portal
Keywords
Details
- Category:
- Monographic publication
- Type:
- rozdział, artykuł w książce - dziele zbiorowym /podręczniku w języku o zasięgu międzynarodowym
- Title of issue:
- Multimedia Communications, Services and Security strony 98 - 109
- Language:
- English
- Publication year:
- 2017
- Bibliographic description:
- Lisowski K., Czyżewski A.: A Method of Object Re-identiciation Applicable to Multicamera Surveillance Systems// / ed. Dziech A., Czyżewski A. eds. Cham: Springer, 2017, s.98-109
- DOI:
- Digital Object Identifier (open in new tab) 10.1007/978-3-319-69911-0_8
- Verified by:
- Gdańsk University of Technology
seen 50 times