以視覺為基礎之停車輔助系統

本論文旨在研究以視覺為基礎之停車輔助系統。以視覺為基礎的停車輔助系統係由空間偵測、路徑規劃與停車控制系統所組成。其中，影像來源係從裝置於車輛側邊的攝影機所取得，序列影像透過一連串的影像處理技術完成停車空間大小的估測。路徑規劃的程序係為了找出合適的路徑以及引導車輛的控制系統自動地調整車輛的轉向。透過不同的影像後置處理技術如：影像變形校正、特徵匹配、逆透視投影。影像變形校正將變形的影像校正成正確的幾何形狀，特徵匹配分析出兩張影像之間的特徵，評估出三維特徵點，使用逆透視投影的標定找出停車空間的深度。另外使用蛇行繪圖來繪製網格建立三維停車空間模型，可以讓駕駛者透過汽車儀表板獲得更多的影像資訊。停車模擬系統使用Matlab/Simulink軟體來完成自動路邊停車的任務。

關鍵字

影像處理；影像變形校正；特徵匹配；逆透視投影

並列摘要

A study of vision-based parking assistance systems has been conducted in this thesis. A vision-based parking assistance system consists of parking space detection, path planning, and maneuver control. The size of parking space is estimated by analyzing captured images from a mounted camera. Path planning is a procedure to find a suitable path as a guide so that the control system can take over the steering control to maneuver the car automatically. Different image processing techniques, such as image distortion correction, feature matching, and inverse perspective mapping, used for parking space detection are explored. In addition, the serpentine drawing method is used to create a 3D display model for drivers to obtain more informative images on dashboard. Simulations of a fuzzy control model using Matlab/Simulink software for car maneuver are also provided in this study.

並列關鍵字

Image processing ； Image distortion correction ； Feature matching ； Inverse perspective mapping

參考文獻

[1] E. Gronenschild and J. Roos, “The accuracy and reproducibility of an algorithm to correct for geometric image distortion in quantitative coronary angiography The Accuracy and Reproducibility of an algorithm to correct for geometric image distortion in quantitative coronary angiography,” IEEE Computers in Cardiology, Sep. 1996, pp.105- 108.

[2] A. M. Muad, A. Hussain, S. A. Samad, M. Marzuki and B. Y. Majlis, “Implementation of inverse perspective mapping algorithm for the development of an automatic lane tracking system,” TENCON 2004, Vol. 1, Nov. 2004, pp.207- 210.

[3] H. G. Jung, D. S. Kim, P. J. Yoon and J. Kim, “Parking Slot Markings Recognition for Automatic Parking Assist System,” IEEE Intelligent Vehicles Symposium, June 13-15, 2006, pp.106-113.

[4] H G. Jung, D. S. Kim, P. J. Yoon and J. Kim, “Light stripe projection based parking space detection for intelligent parking assist system,” IEEE Intelligent Vehicles Symposium, June 13-15, 2007, pp. 962-968.

[5] David G. Lowe, “Distinctive image features form scale-invariant keypoint,” International Journal of Computer Vision, vol.60, no. 4, Nov. 2004, pp. 91~110.

國際替代計量

以視覺為基礎之停車輔助系統

全文下載

主題瀏覽