A method for identifying specific vehicles using template matching

A countermeasure for relieving traffic congestion is to employ a priority signal control system at the intersection, and in order to effect this signal control, it is essential to identifying public transport vehicles. As for the methods for identifying vehicles, while an image processing method for recognizing vehicles by reading the license plate numbers has already been put into practical use, there is no precedence of its application to priority signal control. Therefore, focusing on priority signal control of route buses, we proposed a method for identifying the individual route numbers of the route buses. Considering the ease of processing, we based this method on a template matching process that targets route numbers. However, with this method, the correlation values fluctuate greatly due to the changes in the external environment, which poses the problem of greatly reducing the judgment precision if simple threshold values are used. This paper describes the identification method using the above-mentioned template matching process and its problems, and proposes an identification method that uses a new recognition algorithm for solving those problems. Furthermore, to verify the recognition precision of this method, we carried out an application on an actual road, which resulted in an accuracy rate of 99.6%, which is sufficient for practical use.