License Plate Recognition Using Associative Matching Based on Clique-Finding

Meng Shi, Wataru Ohyama, Tetsushi Wakabayashi, Fumitaka Kimura
2000 IAPR International Workshop on Machine Vision Applications  
In this paper, we propose a new method to separate tlie characters on tlic license plate using associative matching based on clique-finding. 180 digital ixnages are tested by the prototype system. The correct separating rate for each kind of characters have been increased significantly. The experimental results also sliow that this niethod is not only robust against zooming, skewing, variety of illliniination and noise, but also effective to deal with Japanese Hiraganas, Chinese cllaracters and
more » ... srriall numerals, which are very difficult to be obtained by the convexit ional sequential i~rlage segment ation a net hod.
dblp:conf/mva/ShiOWK00 fatcat:vzjofn7kzzaetba2reytb3evom