影像穩定化技術在智慧型運輸系統的應用
No Thumbnail Available
Date
2002-04-??
Journal Title
Journal ISSN
Volume Title
Publisher
國立臺灣師範大學研究發展處
Office of Research and Development
Office of Research and Development
Abstract
本論文針對影像視覺在智慧型運輸系統的應用，提出影像穩定化(image stabilization) 的技術。本研究主要目的在補償固定型交通監控系統之攝影器材因晃動而造成不穩的影像序列。在固定型影像穩定化技術中，可以分成三步驟：運動的估算(motion estimation)，總體運動的分析(global motion analysis)及影像的補償(image compensation)。我們提出k 階投射比對(k-order projection matching)方法，結合方塊比對(block matching)來作運動的估算，前者可以解決當影像含有小角度旋轉的時候，仍可精確地估測運動速度；在第二步驟中我們利用仿射運動模組(affine motion model)來作整體運動的分析，以解決影像含有位移及旋轉的情形；最後利用運動模組的反轉換來還原不穩的影像序列。利用模擬影像及真實影像的實驗結果皆顯示本文所提的技術之可應用性。
In this paper, we propose an image stabilization technique for stabilizing image sequences taken from traffic monitoring cameras. The proposed technique includes three steps: motion estimation, global motion analysis, and image compensation. In the motion estimation, a module combining block matching and k-order projection matching is evoked to estimate motions between two successive images. The process will estimate motions accurately even though there exists rotations between images. In the global motion analysis, an affine motion model is utilized to determine the parameters of translation and rotation between images. The determined affine transformation is then exploited to compensate for the unstabilities of input image sequences. The experimental results of using both simulated and real images have revealed the applicability of the proposed technique.
In this paper, we propose an image stabilization technique for stabilizing image sequences taken from traffic monitoring cameras. The proposed technique includes three steps: motion estimation, global motion analysis, and image compensation. In the motion estimation, a module combining block matching and k-order projection matching is evoked to estimate motions between two successive images. The process will estimate motions accurately even though there exists rotations between images. In the global motion analysis, an affine motion model is utilized to determine the parameters of translation and rotation between images. The determined affine transformation is then exploited to compensate for the unstabilities of input image sequences. The experimental results of using both simulated and real images have revealed the applicability of the proposed technique.