License Plate Localization using Bimodal Attributes

Chirag N. Paunwala and Suprava Patnaik


License plate localization, Salient rules, Bimodal attribute, Diffusion analysis , Connected region analysis, Statistical inconsistency


This paper addresses an intrinsic rule-based LPL algorithm. Foundation takes into account the LP gray features, combined with color standards mandatory for vehicles of a country and hence the title assigned is bimodal attributes. Key contribution is assigning image inferred weights to the rules leading to adaptability in selecting saliency feature. Significant features then over rules other trivial features and the collective measure decides the LPL. Saliency of rules is inherent to the frame under consideration hence all unenthusiastic effects present in the capturing process of the frame are nullified, incorporating great deal of flexibility and more generalization. The probable regions obtained using rule saliency, are then passed through HVS color filtered to confirm the assigned LP colors, which further reduce the False Detection Rate (FDR). In order to claim that the algorithm is better generalized various conditions considered are, variations in illumination, skewed plates, aspect ratio and hence the LP font size, vehicle size, pose, partial occlusion of vehicles and presence of multiple plates. Proposed method allows parallel computation of rules, hence suitable for real time application. The mixed data set has MR = 3.16% and FDR = 5.85%.

Important Links:

Go Back