Vehicle number plate detection in outdoor images
Robustness and efficiency are paramount for number plate detection and recognition systems to ensure their practicality for current day applications. This paper presents algorithms that determine the location of a number plate within still images, with the focus on robustness and efficiency. The number plate locating algorithms exploit the number plate’s signature in order to determine its bounding region. Complex images that contain a significant amount of irrelevant background data were used ensuring the robustness of algorithms developed. Efficiency was gained by removing irrelevant information prior to searching for the number plate signature through the use of pre-processing steps. (a) For the covering entry of this conference, please see ITRD abstract no. E212640.
-
Availability:
- Find a library where document is available. Order URL: http://worldcat.org/isbn/0908556799
-
Authors:
- STEPHENS, M A
- VENNIX, S J
- LAW, R
- HALGAMUGE, S
- Publication Date: 2005
Language
- English
Media Info
- Pagination: 6P
- Monograph Title: Intelligent Vehicles and Road Infrastructure Conference, 16th and 17th February 2005, Melbourne, Australia: Conference proceedings
-
Serial:
- INTELLIGENT VEHICLES AND ROAD INFRASTRUCTURE CONFERENCE, 2005, MELBOURNE, VICTORIA, AUSTRALIA
- Publisher: SOCIETY OF AUTOMOTIVE ENGINEERS, AUSTRALASIA
Subject/Index Terms
- TRT Terms: Automatic vehicle identification; Conferences; Detectors; Image processing; Intelligent transportation systems; License plates; Mathematical models; Vehicles
- ITRD Terms: 8767: Automatic vehicle identification; 8525: Conference; 8633: Image processing; 8735: Intelligent transport system; 6473: Mathematical model; 1346: Number plate; 6120: Sensor; 1255: Vehicle
- Subject Areas: Operations and Traffic Management; Vehicles and Equipment; I71: Traffic Theory; I73: Traffic Control;
Filing Info
- Accession Number: 01014605
- Record Type: Publication
- Source Agency: ARRB
- ISBN: 0908556799
- Files: ITRD, ATRI
- Created Date: Dec 22 2005 3:05PM