以分界特徵限制為基礎之即時道路邊緣追蹤演算法

A New Method of Efficient Road Following Algorithm Based on Temporal Region Ratio and Edge Constraint

陳宣輯(Shen-Chi Chen)

指導教授 ： 林進燈

國立交通大學/資訊學院/生醫工程研究所/碩士(2009年)

https://doi.org/10.6842/NCTU.2009.00818

摘要

近年來隨著車輛數目的增加與性能不斷的成長，道路情況與駕車難度也相對的提升，當今的車輛仍需要駕駛者全神貫注地人為操作，相對地增加駕駛人的壓力與車禍發生的可能性。隨著行車安全愈來愈受到人們重視，為了改善此問題，世界各國的研究機構、知名車廠、學術研究室均投入大量的精力於智慧型運輸系統ITS(Intelligent Transportation System) 中的安全駕駛輔助系統，並得到許多人 的期待與讚許。也因此這類的先進式車輛控制系統的願景，以從當初的駕駛者安全輔助性質轉為主動式的安全操縱系統，甚至最終的目標已經設定為自動駕駛。 行駛道路邊緣的偵測為其中一項關鍵的技術，本論文將以CCD (charge coupled device)攝影機所擷取的影像為基礎利用本論文提出的邊緣特徵穩定地實現道路邊緣偵測與追蹤的演算法。本研究所提出的道路邊緣特徵結合了邊緣強度與道路顏色兩項重要的特徵，改進了當今許多研究只利用其中一類可能會產生的問題。我們將在論文中說明如何利用所提出的創新演算法修正其他文獻的問題。本研究的演算法利用有效的特徵，增加系統之適應性，可處理多類型的道路邊緣，包括一般市區道路、高速公路、等有明顯車道線的邊緣；另外，由於我們所採用的邊界特徵擁有突出的穩定性與一致性；因此本系統對於一些鄉間小路、產業道路..等，沒有明顯的標線邊緣，與夜間道路仍能提供有效性地偵測結果。本系統在設計時在考慮許多駕駛狀況，因此，系統具有處理一般駕駛狀況與道路環境的改變的能力，如變換車道，左右偏動，與不平坦道路，造成相機的晃動。本篇論文提出一主車道邊緣偵側與追蹤系統，其具有先進的適應性與即時的處理效率。

關鍵字

影像式 ； 道路追蹤 ； 主車道邊緣偵測

並列摘要

In recent years, since the number and properties of the vehicle are continuous increasing, the complexity of road condition and the difficulties of operation increase relatively. As of today, physically driving not only requires drivers’ full attention but could potentially amplify drivers’ stress and possibility of car accidents. As people become more concerned about the safety in driving, many research institutes around the world, well-known automobile manufactures, and academic research labs invest a tremendous amount of effort to develop a safe driving assistance system of ITS (Intelligent Transportation System) to alleviate the problems of physical driving. With many expectation and applaud, the advanced supporting system has transformed into an active safe operating system, and ultimately will develop into an automatic driving system. Since the driving path boundary detection is one of the key techniques of automatic driving system, this paper is based on the use of images captured by the CCD (charge coupled device) camera and we discovered the general boundary features to steadily achieve main driving path boundary detection and tracking system.
We purpose to integrate two major features, edge intensity and color distribution, of road boundary and mitigate the problems encountered by many other studies which utilize only one major feature for road boundary detection. In this paper we will explain how to make use of the advanced algorithms to amend the problems in other literatures. Our algorithms utilize effective features to enhance the adaptability of the system; the system is able to manage various types of road boundary includes general urban road, highway, or any boundary with the clear lane line. In addition, we use the general characteristics of the boundary, and therefore the system is still able to effectively provide detection result at the road without clear boundary such as the countryside paths, the road with no lane line, and the scene at the night. In the design of the system we take into account many driving situations, such as changing lanes, left and right side move, and non-flat roads lead to camera shake. Therefore, the system possesses good ability to deal with general driving conditions and variations. This paper presents a system which includes main driving path boundary detection and boundary tracking with advanced real-time adaptability and efficiency.

並列關鍵字

vision-based ； road following ； driving path boundary detection

參考文獻

[1] J. Crisman and C. Thorpe, SCARF:A Color Vision System that Tracks Roads and Intersections, Conf. Proc. IEEE International Conf. on Robotics and Automation, IEEE, 1993

[2] J. Crisman and C. Thorpe, UNSCARF: A Color Vision System for the Detection of Unstructured Roads, Conf. Proc. IEEE International Conf. on Robotics and Automation, 1991.

[3] J.Huang and B Kong, A New Method of Unstructured Road Detection Based on HSV Color Space and Road Features, International Conf. on Information Acquisition, 2007.

[4] Y.Alon, A.Ferencz and A.Shashua, Off-road path following using region classification and geometric projection constrains, Computer Vision and Pattern Recognition, IEEE Computer Soc. Conf, 2006.

[5] C.Rasmussen, Grouping dominant orientations for ill-structured road following, Proc. IEEE Soc. Conf. Computer Vision and Pattern Recognition, 2004.