An improved intelligent transportation system: An approach for bilingual license plate recognition

Abstract

An intelligent transportation system (ITS) is the leading-edge technology that is used to control the traffic and prevent the rule violators. The ITS systems are equipped with an automatic number plate recognition (ANPR) techniques which recognize the license plate of the vehicles. This paper proposes an approach for Hindi license plate recognition and extends the ability of ITS systems. In the proposed approach, connected component labeling is used for character segmentation and histograms of oriented gradient (HOG) features are used to classify the Hindi characters. We propose an integrated classification model for character classification. To classify the alphabets and numerals, a ANN model is designed and trained for datasets. The approach segments the characters with 100% accuracy, and average accuracy of the classification model is 96.7%.