高速攝影機無論是在工業界、學術界以及其它領域皆被廣泛的應用，但一直以來，其昂貴的價格居高不下，也是普遍使用者較不能接受的，基於這個出發點，我們的目的是希望能開發出低價位的高速攝影系統。其實在2001年美國史丹佛大學的計算機圖形實驗室曾開發出「史丹佛相機陣列」的架構[1]，不過該系統之硬體架構過於龐大，在使用上會有所限制，甚至將來實現產品化也有困難，因此我們決定導入嵌入式系統的概念，改善史丹佛相機陣列的高速攝影系統，用更便宜以及較易於攜帶與使用的系統平台來實現。
在論文[3][4]中，我們已經成功的以SoPC平台開發出相機陣列的高速攝影系統。由於平台周邊與系統資源限制的關係，當時只能以兩顆CMOS相機實現，最快可達120fps的拍攝速度。為了能進一步的提高拍攝速度與系統效能，我們將持續致力於此架構下的軟硬體系統開發。在本論文中，我們將提高攝影系統的拍攝速度，連接多4倍的相機數目8顆CMOS cameras以達到每秒480張的拍攝數度。另外，我們修改了自動曝光和位移誤差演算法，減少執行曝光的次數達31.17%，以及降低位移誤差的複雜度，可有效改善系統的執行速度。

The High speed video camera is generally used in industry, academia to study high speed event. But expensive for mostly user, our purpose is cost down the high speed video camera system. In fact, the Stanford Computer Graphic Laboratory has developed a “Stanford Multiple Camera Array” architecture and provided a good result in 2001 [1]. However, the scale of the hardware architecture in their design is too large to be implemented as a portable device. Therefore, we decided to realize the system by reducing the architecture with an embedded platform.
On this concept, we have redesigned the high speed video camera system based on SoPC platform in [3][4]. Previous effort of our lab has achieved 120fps frame rate. In order to improve frame rate and system performance further, we keep developing the system. In this paper, we has 8 CMOS cameras in our system and frame rate could achieve 480fps. To improve the system execution time, we modify the auto-exposure algorithm to reduce the counts of exposure up to 31.17% and alter the displacement error rectification algorithm to diminish the complexity of displacement.

[1] Bennett Wilburn, E. V. Talvala, Emilio Antunez, Mark Horowitz, Neel Joshi,
Vaibhav Vaish, Adam Barth, Andrew Adams, Marc Levoy, “High Performance
Imaging Using Large Camera Arrays,” In ACM Trans. on Graphics, vol. 24,
no. 3, pp. 765-776, July 2005.
[2] Wikipedia, “Eadweard Muybridge,”
http://en.wikipedia.org/wiki/Eadweard_Muybridge
[3] 鄭子凡，“以多組CMOS感測器實現高速攝影系統之SOPC硬體架構”，國立台灣科技大
學電子工程系碩士論文，2007
[4] 陳首元，”基於SOPC平台之相機陣列式高速攝影系統軟體架構及影像校正設計”， 國
立台灣科技大學電子工程系碩士論文，2007
[5] Altera, “DE2-70 Development and Education Board,”
http://www.edastech.net/de270.html
[6] B. R. Lim, R. H. Park, “High Dynamic Range for Contrast Enhancement,”
IEEE Trans. On Consumer Electronics, vol. 52, no. 4. pp. 1454-1462, Nov.
2006.
[7] Kodak, “Shutter Operations for CCD and CMOS Image Sensors,” Kodak Image
Sensor solutions, Revision 2.0, 2003.
[8] J. Y. Liang, Y. J. Qin and Z. L. Hong, “An Auto-exposure Algorithm for
Detecting High Contrast Lighting Conditions,” Proc. Of the 7th Int. Conf.
on ASIC, Guilin, Peoples R. China, vols. 1 and 2, pp. 725-728, Oct. 2007.
[9] 數位影像坊, ”Mr. OH! 數位攝影講座,”
http://www.digital.idv.tw/DIGITAL/Classroom/MROH-CLASS/index-classroom-
MROH.htm
[10]H. Longuet-Higgins, “A computer algorithm for reconstructing a scene from
two projections,” Nature, Vol. 293, no. 10, pp.133-135, 1981.
[11]Papadimitriou, D.V. and Dennis, T.J., “Epipolar line estimation and
rectification for stereo image pairs,” IEEE Transactions On Image
Process, 5(4), pp. 672-676, April 1996.
[12]Hartley, R. and Zisserman,.A., “Multiple View Geometry in Computer
Vision,” Cambridge University Press, 2004.
[13]Pollefeys M, Koch R, van Gool L. “A simple and efficient rectification
method for general motion.” In Werner B, ed. Proc. of the Int'l Conf. on
Computer Vision. Los Alamitos: IEEE Press, 1999. 496~501.