Bilateral partitioning based character recognition for vehicle license plate

Abstract

This paper presents a new methodology for the image segmentation and character recognition from standard Indian License number plates. This method first, gets input of the segmented characters that is partitioned by bilateral partitioning method, in which we eliminate similar part from the character and match it by judging template and return identified character. This partitioning may be performed horizontally or vertically. The characters are to be partitioned horizontally or vertically depend on their subgroup, and before this subgroup the characters are grouped on the basis of the number of holes in it. The subgroup is made on the basis of some similar features like | , / , \ , - , ( , - , etc. Suppose we have alphabet T and I here similar portion is I than both will go to same subgroup and we partition it horizontally. This method eliminates the problem of confusion between similar looking elements like C, G and T, I, 1, J etc by exploiting the small but significant differences among them. © 2011 Springer-Verlag.