초록

용어보기논문에서 용어와 풀이말을 자동 추출한 결과로, 시범 서비스 중입니다. AI-Helper

Abstract ▼ AI-Helper

In this paper proposes using mono camera for real-time lane detection, distance estimation and object detection system based on deep learning network. After removing the perspective of the acquired image data, the candidate group corresponding to the lane is selected using the Sliding Windows method...

In this paper proposes using mono camera for real-time lane detection, distance estimation and object detection system based on deep learning network. After removing the perspective of the acquired image data, the candidate group corresponding to the lane is selected using the Sliding Windows method, and the lane is accurately detected through the RANSAC technique. For object detection, we use a deep learning algorithm that has a higher detection rate than the existing computer vision algorithm and robust to changes in the surrounding environment. Based on the reference data, the formula is established by using the relation between the Bounding Box Pixel value of the object detected by the monocular camera and the actual distance information, Then, the distance between the object and the vehicle is estimated using the formula. And for the stable driving of autonomous vehicle, our autonomous driving system are used longitudinal and lateral control. Therefore, in this paper, the performance of the proposed system is verified by experiment of lane keeping and risk discrimination and Emergency Braking experiment according to distance from forward object.

주제어