Indonesian license plate recognition with improved horizontal-vertical edge projection

<p>License plate recognition (LPR) is one of the classical problems in field object recognition. Its application very crucial automation transportation system since it helps to recognise a vehicle identity, which information stored license plate. LPR usually consists three major phases: pre-processing, localisation, optical character (OCR). Despite being classical, its implementation faced with much more complicated real scenario. This paper proposed an improved algorithm based on modified horizontal-vertical edge Projection. The method uses for detecting and localising region interest. It done using horizontal vertical projection image. Related works proved that simplest be implemented, yet effective against Indonesian However, performance gets reduced when specular reflection occurs sample Therefore, morphological operations are utilised pre-processing phase reduce such effects while preserving needed information. Eighty images captured various camera configurations were used this research. Based experimental results, our shows improvement compared previous study successfully detect 71 plates 80 image samples results 88.75% accuracy.</p>