Autonomous reverse parking system - vision approach through ridge detector and Kalman filter

You are here: Homepage > Search

Autonomous reverse parking system - vision approach through ridge detector and Kalman filter

Du, Xinxin / Tan, Kok Kiong / Htet, Kyaw Ko Ko

Some commercialised vehicle models have been equipped with semi-autonomous parking systems. However, gaps to fully automated solutions still exist and cost considerations further constraint its acceptance among consumers. This paper proposes a low-cost vision-based approach to fully self-reverse parking system. The vision is based on ridge detector and Kalman filter to identify and track the parking slot boundaries. The vehicle position and orientation are estimated from the vision system in real-time. It serves as a consistent feedback to the revised sliding mode control (SMC) system which controls the vehicle to follow the desired path. Due to the high estimation accuracy and consistency from the vision system, the results obtained based on 200 on-field tests illustrated that the proposed approach was able to park the vehicle robustly. With its easy setup and excellent performance, this approach can be implemented onto existing vehicles practically with minimal additional cost.