Title :
Influence of driver behavior patterns in correcting video sensing errors in traffic surveillance applications
Author :
Naghiu, Florica ; Pescaru, Dan
Author_Institution :
Dept. of Comput., Politeh. Univ. of Timisoara, Timisoara, Romania
Abstract :
Correcting video sensing error problems for traffic surveillance was hard investigated over the last years. Several parameters such as light intensity, image quality and resolution, and in general all the key factors related to image capture were considered. We have investigated here a modified MCMC approach based on the influence of typical patterns of driving behavior. These patterns are tested and compared against the standard MCMC algorithm. Results prove achievements in error reduction.
Keywords :
driver information systems; filtering theory; probability; video surveillance; MCMC approach; Markov Chain Monte Carlo method; driver behavior pattern; filtering theory; image quality; image resolution; light intensity; probability; traffic surveillance application; video sensing error problem; Bayesian methods; Cameras; Computer errors; Error correction; Monte Carlo methods; Particle filters; State estimation; Surveillance; Traffic control; Vehicles;
Conference_Titel :
Intelligent Computer Communication and Processing, 2009. ICCP 2009. IEEE 5th International Conference on
Conference_Location :
Cluj-Napoca
Print_ISBN :
978-1-4244-5007-7
DOI :
10.1109/ICCP.2009.5284764
