초록 열기/닫기 버튼
Presented in this paper is a method of searching the parking space for an electric vehicle, and of tracking a wireless charging spot accurately from the bird’s-eye-view image provided by the AVM(around view monitoring) system, by fusing four fish-eye cameras already mounted on produced vehicles. In addition, how to get parameters and combine four fish-eye cameras for AVM images throughout camera calibration is introduced. The proposed detection method finds parking spaces based on vision using image segmentation. Assembling enormous datasets can be a very daunting and time-consuming task due to the manual effort of collecting and labeling data. To solve this problem, this paper used data augmentation, a data space solution to the problem of limited data. The proposed tracking method is robust against problems like dim lighting, background, and reflections on road surfaces.
키워드열기/닫기 버튼
Around view monitoring system, Camera calibration, Data augmentation, Parking slot detection, Bird’s eye view image
