Design and Evaluation of Parking Position Detection with Human Cooperation for Automatic Parking

Abstract

Whenever we ride in an automobile, we must park it. Then, we want to avoid that very troublesome act if possible because it requires a certain amount of driving skills. In recent years, the parking assist system is becoming partially implementation. This conventional system mainly recognizes the parking space surrounded by the lane. However, this conventional system is merely driving support system for supporting safety driving corresponding to the level of a driving technique of a driver. The automatic parking system have been studied previously. However, that system have an issue to depend on development of infrastructure at parking lots to detect the parking space, such as cameras and sensors or the communication technology. Then, the authors suggest the new automatic parking system that drivers indicate parking position from the outside a car and assist the environmental recognition of the car to allow a car to park semi-automatically. This new system is another approach to park an automobile with human cooperation. Specifically, the driver parks using an onboard camera and the indication device which indicate parking position. An onboard camera detects the indication device, therefore a vehicle detect parking position. The method to detect indication device is based on HSV-colored conversion and Hough transformation. The method to calculate between the parking position and the vehicle is based on onboard camera. Assuming some real parking positions, we evaluated making the vehicle detect parking positions using the method that we proposed. These results show enough detection to park possible and possibility of the proposed system implementation.