Matching RFID and image‐based vehicle identification data to detect illegal road use

Abstract

Effective road law enforcement is a vital part of transportation systems. Escalating traffic congestion and widespread illegal road usage pose fresh hurdles for law enforcement systems. An intelligent law enforcement system that aims to reduce illegal road use must enable immediate intervention on a tactical level without disrupting the flow of legal vehicles. To achieve a fast tactical response, strategic-level information, such as location and identity, must ensure law enforcement officials are in the right place at the right time. This thesis presents a system that equips authorities with the correct information to act against illegal road users while maintaining normal traffic flow. The proposed system improves existing solutions by combining Radio Frequency Identification (RFID) and Computer Vision (CV) data to detect and identify illegal vehicles. The authorities can leverage this information to remove such vehicles from the road without affecting the legal road user's experience. This thesis will investigate the performance criteria for such a system within a real-world environment and will determine which combination of data fields stored on the license plate tag can enable accurate vehicle identification while still being readable within a practical setup. Practical measurements will verify if the theoretical requirements can be achieved in practice.