利用強健的車道標線候選點選取方法實現在複雜環境下之車道線偵測

Abstract

近年來行車輔助系統DAS(Driver Assistance Systems)越來越盛行，例如車道偏移警示、停車輔助系統、自動剎車系統...，其中大多數的方法都是需要用到車道線偵測的技術，駕駛需要使用到行車輔助系統大多是出現在下大雨或市區道路時，因此我們開發了一個強健的車道線偵測與穩定的車道線追蹤系統。在實際應用上它會在很多常見的場景下發生問題，例如 路面標線被遮蔽、有影子在影像中出現、光線強弱的變化(白天、晚上、有雲遮住光線、進出隧道…)、影像模糊(如下雨、下雪、起霧…) 等狀況。在我們的研究裡，我們透過學習法來解決這些問題，首先我們的輸入影像是車前的影像(行車紀錄器)，以下我們提出了三個方法：1) 基於俯視影像的車道標線候選點選取方法 2) 一個對於惡劣環境的偵測加強方法 3) 一個在沒有車速資訊下的車道線追蹤，其中在面對車道標線候選點選取方法的部份我們選擇了幾種分類器來實現他，經過比較後我們選用了AdaBoost分類器，最後我們驗證所提出來的方法是可以成功的解決上述在實際場景下會發生的挑戰，而且我們的方法可以在分別在640x480與320x240的影像輸入狀況下達到每秒17張與65張的運算速度。

In recent years Driving Assistance Systems DAS (Driver Assistance Systems), such as lane departure warning, parking assist system, and automatic braking systems ..., has become more and more popular. Most of the methods are needed to use lane detection techniques, and the driving assistance systems are needed to use mostly appear in the next heavy rain or urban roads. Therefore we have developed a robust lane detection and lane tracking system. It is a hard problem primarily due to large appearance variations in lane markings caused by factors such as occlusion (traffic on the road), shadows (from objects like trees), and changing lighting conditions of the scene (transition from day to night). In this paper, we address these issues through a learning-based approach using visual inputs from a camera mounted in front of a vehicle. We propose the following: 1) a lane marking candidates selection method; 2) a rule-based enhancement to help the detection in challenging scenarios; and 3) a tracking rule to track the lane markings, without knowledge of vehicle speed. In the lane marking candidates selection method, we have chosen several classifiers to implement and by comparison we choose the AdaBoost classifier. We investigate the effectiveness of our algorithm on challenging daylight, night-time, heavy rain and urban street scene video sequences. Furthermore, our method can work at FPS 17 at 640x480 input images and FPS 65 at 320x240 input images.