Visual Path Tracking Control for Park Scene

Abstract

Autonomous driving application is developing towards specific scenes. Park scene has features such as low speed, fixed routes, short connection, less complex traffic, and hence is suitable for bringing autonomous driving technology into reality. This paper targets park scene, and proposes a visual path tracking lateral control method using only one webcam. First, we calculate error of distance and error of angle from camera images, and then use fuzzy logic to fuzzify them into a combined error degree. The PID control algorithm takes it as input, and outputs steering wheel angle control command. Fuzzification could tolerate the error brought by image transformation and lane detection, making PID control more stably. Our experiments in both virtual and real scene show that our method can accurately and robustly follow the path, even at night. Compared with pure pursuit, our method can make 5 meters turning.Autonomous driving application requires safe and robust technology with reasonable cost. More importantly, we believe the self-driving application scenario should be redesigned for it to work. In a develop zone park, we redesign the short connection self-driving scenario, and manage to run trial operation. In our redesign, lane markers of routes are painted on the road, besides two lanes markers, we also add another lane marker in the center of road. People are not allowed to enter the road. The route is fixed, with a number of stops for passengers to get onto the self-driving car. Passengers order cars through an APP online, and are expected to arrive reserved stops on time. Then passengers scan the QR code to start their journey. This paper focuses on the visual path tracking control method applied to our short connection self-driving application, and we will discuss our redesigned self-driving application in another paper. It has merits such as accurately and robust path following, path following at night, small radius turning, all with reasonable cost.