TY - GEN
T1 - Surveillance for security as a pursuit-evasion game
AU - Bhattacharya, Sourabh
AU - Başar, Tamer
AU - Falcone, Maurizio
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2014.
PY - 2014
Y1 - 2014
N2 - This work addresses a visibility-based target tracking problem that arises in autonomous surveillance for covert security applications. Consider a mobile observer, equipped with a camera, tracking a target in an environment containing obstacles. The interaction between the target and the observer is assumed to be adversarial in order to obtain control strategies for the observer that guarantee some tracking performance. Due to the presence of obstacles, this problem is formulated as a game with state constraints. Based on our previous work in [6] which shows the existence of a value function, we present an off-line solution to the problem of computing the value function using a Fast Marching Semi- Lagrangian numerical scheme, originally presented in [15]. Then we obtain the optimal trajectories for both players, and compare the performance of the current scheme with the Fully Discrete Semi-Lagrangian Scheme presented in [6] based on simulation results.
AB - This work addresses a visibility-based target tracking problem that arises in autonomous surveillance for covert security applications. Consider a mobile observer, equipped with a camera, tracking a target in an environment containing obstacles. The interaction between the target and the observer is assumed to be adversarial in order to obtain control strategies for the observer that guarantee some tracking performance. Due to the presence of obstacles, this problem is formulated as a game with state constraints. Based on our previous work in [6] which shows the existence of a value function, we present an off-line solution to the problem of computing the value function using a Fast Marching Semi- Lagrangian numerical scheme, originally presented in [15]. Then we obtain the optimal trajectories for both players, and compare the performance of the current scheme with the Fully Discrete Semi-Lagrangian Scheme presented in [6] based on simulation results.
KW - Fast marching
KW - Pursuit-evasion games
KW - Semi-Lagrangian schemes
UR - http://www.scopus.com/inward/record.url?scp=84910008267&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84910008267&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-12601-2_23
DO - 10.1007/978-3-319-12601-2_23
M3 - Conference contribution
AN - SCOPUS:84910008267
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 370
EP - 379
BT - Decision and GameTheory for Security - 5th International Conference, GameSec 2014, Proceedings
A2 - Poovendran, Radha
A2 - Saad, Walid
PB - Springer
T2 - 5th International Conference on Decision and GameTheory for Security, GameSec 2014
Y2 - 6 November 2014 through 7 November 2014
ER -