Automatic Number Plate Recognition Application for Metropolitan Toll Road Payment System

Abstract

Many factors cause congestion on toll roads, one of which is service time at toll gates. Maximum service times for each vehicle be regulated by Indonesia Toll Authority to no more than 6 seconds for the open transaction toll system. There-fore, an electronic toll road payment system integrated with ANPR (Automatic Number Plate Recognition) technology is proposed. The proposed system can identify vehicle plates automatically to make payments. The Optical Character Recognition (OCR) method was chosen in this study for recognizing the number plate. ANPR with Tesseract OCR is placed inside the computer. Based on the results of the analysis and testing that has been carried out, the low average accuracy of 63,14% from this research is small due to the small number of sample images. The average execution times of 1405 milliseconds from this research are be-low the maximum 6 seconds limit for open transaction toll collection in Indonesia. One failed plate localization from sample images has founded in a car image with number plate DK 1547 EC.