DocumentCode
2366803
Title
Highway traffic information extraction from Skycam MPEG video
Author
Yu, Xiao-Dong ; Ling-Yu Duan ; Tian, Qi
Author_Institution
Labs. for Inf. Technol., Nat. Univ. of Singapore, Singapore
fYear
2002
fDate
2002
Firstpage
37
Lastpage
42
Abstract
Skycams are video cameras that are mounted at high altitudes to provide images of large areas. We present an algorithm that quickly extracts highway traffic information, such as average vehicle speed and density, directly from MPEG compressed Skycam video. Our experiments show that 80-90% of estimation accuracy can be achieved for both vehicle speed and density. The algorithm is stable and effective for Skycam, which the traditional approaches cannot handle due to the low resolution. Because a Skycam can cover a much longer stretch of road than a close-view camera that is used in the current traffic monitoring system, great savings can be expected by widely deploying a Skycam based traffic monitoring system.
Keywords
data compression; discrete cosine transforms; image texture; monitoring; motion compensation; road traffic; surveillance; video cameras; video coding; Skycam MPEG video; average vehicle density; average vehicle speed; estimation accuracy; highway traffic information extraction; traffic surveillance; video cameras; Cameras; Data mining; Digital images; Information technology; Laboratories; Monitoring; Road transportation; Shape; Surveillance; Vehicle detection;
fLanguage
English
Publisher
ieee
Conference_Titel
Intelligent Transportation Systems, 2002. Proceedings. The IEEE 5th International Conference on
Print_ISBN
0-7803-7389-8
Type
conf
DOI
10.1109/ITSC.2002.1041185
Filename
1041185
Link To Document