차선 추정 정확도 향상을 위한 알고리즘 개발

Abstract

To develop advanced driver assistance systems (ADAS), the accuracy of lane detection from the camera sensor is very important. However, if the camera sensor is only used, there is a possibility that the lane cannot be recognized in a situation where the lane is blurry or blocked by a front vehicle, and the weather is inclement. In order to prevent these phenomena, various methods of correcting the undetectable or erroneous lane detection through the camera sensor have been researched, such as using DGPS or vehicle dynamic model. In this paper, we propose a lane estimation method based on two different vehicle model that fuses both camera sensor and vehicle IMU information. The proposed lane estimation algorithm is verified in real vehicle tested data using the computer software MATLAB/Simulink.