VIL을 위한 비전센서 최적화 모델링

Abstract

Recently, the field of ITS ( intelligent transportation systems) has become more important in the automobile industry, and ADAS (Advanced Driver Assistance System) is actively developed for the safety and convenience of drivers. Vehicle mounting of ADAS technology is becoming mandatory as a legal requirement. Verification of real vehicle performanceis difficult because of the time, cost and risk. Therefore, the VIL (Vehicle In Loop) method based on the virtual driving environment is applied to verify the ADAS function. In this paper, we sampled the detected lane data in the road database. Sampling is performed based on the car position setting and the vision sensor model. In order to optimize the search speed, the lane search is divided into the first search and the subsequent search in the lane search.