Title :
Real-time license plate identification by perceptual shape grouping and tracking
Author :
Chen, Huiqiong ; Rivait, Derek ; Gao, Qigang
Author_Institution :
Fac. of Comput. Sci., Dalhousie Univ., Halifax, NS
Abstract :
This paper presents a perceptual organization based method for real-time license plate identification and tracking by video content analysis. In this method, video content is described using a set of perceptual shape features, called generic edge tokens (GET). A video frame can be represented as a GET map. Motion GETs (MGETs) are segmented from the consecutive images based on GET map and motion clue. A MGET graph is proposed for coding the moving content in video sequence. A license plate is identified by searching a sub-MGET-graph (SMG) that satisfies the license plate shape model. This target shape model is pre-defined by a set of recognition rules according to the GET based shape representation. The SMG representing the license plate can be detected by perceptually grouping the plate shape in MGET graph. The license plate is then tracked on the region of interest (ROI) predicted based on the motion continuity, so that the search can be focused to the most relevant sub-region of the image instead of the entire image. Accordingly, the data flow to be processed is reduced significantly based on perception clues and the motion pattern prediction. This system may be adapted for other target identification tasks by updating a subset of the recognition rules. The efficiency and effectiveness of this method are demonstrated using a gate way setting camera application
Keywords :
automated highways; graph theory; image motion analysis; image recognition; image segmentation; real-time systems; target tracking; traffic engineering computing; video coding; image segmentation; license plate tracking; motion generic edge token graph; motion pattern prediction; moving content coding; perceptual shape grouping; perceptual shape tracking; real-time license plate identification; shape representation; video content analysis; video sequence; Computational modeling; Computer science; Feature extraction; Image edge detection; Information geometry; Intelligent transportation systems; Licenses; Partitioning algorithms; Shape; Vehicles;
Conference_Titel :
Intelligent Transportation Systems Conference, 2006. ITSC '06. IEEE
Conference_Location :
Toronto, Ont.
Print_ISBN :
1-4244-0093-7
Electronic_ISBN :
1-4244-0094-5
DOI :
10.1109/ITSC.2006.1707411