제목

선 분할 모델을 이용한 차선 검출

Curve Lane Fitting Based on Line-Segment Model

저자명

린칭(Lin Qing), 한영준(Youngjoon Han), 한헌수(HernSoo Hahn)

영 문초록

This paper presents a curve lane fitting model to represent curve lane in the image using line segments. The curve lane model is composed of a set of straight line segments that are connected with each other head to tail. In order to find the best line segment sets, lane edges are first extracted and the edge image is then divided into several horizontal sections. Segmented Hough Transform is applied inside each section to vote for a line segment using the edge pixels contained in this section. Finally these line segments are evaluated based on global constraints of lane model and connected with each other from bottom to top sections. Experiment results show that this model can be used to approximate curve lane effectively, and the processing time is about 50~60ms for a single image with resolution of 320×240.

원문 수록처

제어로봇시스템학회 학술대회
Pages 507~509 (3 pages)

자료유형 학술대회자료