A Multitarget Tracking Video System Based on Fuzzy and Neuro-Fuzzy Techniques
- PDF / 3,611,606 Bytes
- 18 Pages / 600 x 792 pts Page_size
- 90 Downloads / 202 Views
A Multitarget Tracking Video System Based on Fuzzy and Neuro-Fuzzy Techniques ´ Garc´ıa Jesus Departamento de Inform´atica, Universidad Carlos III de Madrid, Avda de la Universidad Carlos III 22, Colmenarejo 28270, Spain Email: [email protected]
Jose´ M. Molina Departamento de Inform´atica, Universidad Carlos III de Madrid, Avda de la Universidad Carlos III 22, Colmenarejo 28270, Spain Email: [email protected]
Juan A. Besada E.T.S.I. Telecomunicaci´on, Universidad Polit´ecnica de Madrid, Ciudad Universitaria s/n, Madrid 28040, Spain Email: [email protected]
Javier I. Portillo E.T.S.I. Telecomunicaci´on, Universidad Polit´ecnica de Madrid, Ciudad Universitaria s/n, Madrid 28040, Spain Email: [email protected] Received 19 December 2003; Revised 23 December 2004 Automatic surveillance of airport surface is one of the core components of advanced surface movement, guidance, and control systems (A-SMGCS). This function is in charge of the automatic detection, identification, and tracking of all interesting targets (aircraft and relevant ground vehicles) in the airport movement area. This paper presents a novel approach for object tracking based on sequences of video images. A fuzzy system has been developed to ponder update decisions both for the trajectories and shapes estimated for targets from the image regions extracted in the images. The advantages of this approach are robustness, flexibility in the design to adapt to different situations, and efficiency for operation in real time, avoiding combinatorial enumeration. Results obtained in representative ground operations show the system capabilities to solve complex scenarios and improve tracking accuracy. Finally, an automatic procedure, based on neuro-fuzzy techniques, has been applied in order to obtain a set of rules from representative examples. Validation of learned system shows the capability to learn the suitable tracker decisions. Keywords and phrases: fuzzy-knowledge-based system, neuro-fuzzy learning, video image tracking, data association.
1.
INTRODUCTION
In airport areas, advanced surface movement, guidance, and control systems (A-SMGCS) [10] are conceived as new procedures and technologies to support ground traffic management, increasing both safety and efficiency of traffic flow in complex, high-density airport ground scenarios. One of the core functions within A-SMGCS is surveillance, in charge of the automatic detection and tracking of all relevant targets located in the airport movement area (runways, taxiways, and apron areas). These targets moving in the airport are generally commercial aviation aircraft and surface vehicles, such as fuel trucks, luggage convoys, cars, and so forth. A-SMGCS processes data from different types of sensors
to monitor all ground traffic providing controllers with a periodically updated synthetic image containing all interesting traffic state on the airport surface. In this paper we will focus on tracking aspects when the data to be processed are provided by cameras. They act as noncooperative sensors, so not req
Data Loading...
