交通計画(2)

[IV-152] VEHICLE DETECTION AND COUNTING FOR VEHICLE TYPES OF MOTORCYCLE AND AUTO RICKSHAW USING DEEP LEARNING

〇Tashiro Daichi¹, Tetsuhiro Ishizaka¹ (1.NIHON University)

Keywords:Vehicle Detection, Image Processing, CNN

The accuracy rate of the vehicle counting by the conventional traffic sensor in the Paldi intersection of Ahmedabad city, India, is around 60%. This low accuracy is caused that most vehicle of the auto rickshaw and motorcycle is not detected with high accuracy. Recently, vehicle detection on image processing has been developing and achieved good performance.
Therefore, the object of this study is to verify whether the vehicle detection and counting of motorcycle and auto rickshaw is performed using deep learning for the picture of the CCTV camera in Ahmedabad.