Extraction of Text from Vehicle Number Plate Of a Specific Region Using Template Matching

ABSTRACT The registration number of a vehicle is extracted from the image of the number plate in a text format using morphological processes and template matching for character identification. Due to non-standardization of the number plates in India, many a times there is an interchangeable detection of alphabets and numbers in the extraction process. A region specific algorithm has been implemented, based on the region specific number plates. For instance, in India, the registration number is 10-digit alphanumeric code, with first, second, fifth and sixth character as alphabet and other characters as number. This methodology increases the accuracy of text identification by removing the probability of an alphabet being produced as the result of correlation at the place of a numeric and vice versa