不同於目前的自動骰點辨識系統，僅適用於製造商設計之密閉環境下的骰子機遊戲機台，本研究是第一個針對開放環境下的骰子遊戲桌台而設計的骰點辨識系統。本系統可應用於一般的博奕場所，一般博奕場所內所使用的骰盅與骰子等相關硬體元件，無需更換，僅需加裝本系統所設計之軟體系統和攝影機於適當位置，即可進行自動骰點辨識。
目前的自動骰點辨識系統，均使用單台攝影機來擷取骰子的平面影像進行骰點辨識。本研究使用複數台攝影機擷取不同視角、距離與高度的骰子影像，利用各影像之間的幾何空間資訊，擷取骰子多面向的影像，經電腦立體視覺與人工智慧的分析，取得各骰點分佈的位置，並依據各骰點的分佈模式，判讀各骰子之正確點數，最後透過投票機制辨識各骰點的結果。實驗證明本研究所提出之骰點辨識系統可在多種不同之光照環境下，達到精確的辨識效果。

Different from all existing dice recognition systems which only work for the controlled environments defined by the vendors, this research focuses on dice recognition in an uncontrolled environment, aiming at the applications in generic table games. The primary challenge comes from various possible illumination conditions, which are not considered in all existing dice recognition systems. Existing dice recognition systems use single camera placed on top of the dices, capturing images and recognizing the dots on the dices in a closed and controlled environment by two dimensional image processing methods. This research applies stereo vision with multiple cameras to identify the dices and recognize the dots. The method is composed of six steps: (1) camera calibration to set up the correspondences between different views, (2) segmentation of dice regions, (3) search for dot candidates, (4) design and making of dot classifiers, (5) recognition of dots distribution pattern, and (6) system integration. Experiments show that the proposed stereo vision based system can recognize dices in various illumination conditions, revealing its potential to be applicable in uncontrolled environments and thus the integration with generic table games.

[1] 尊博骰寶, 尊博科技股份有限公司。
[2] Fumihiko Itagaki, Satoshi Murakami, “Die Eye Number Determination Method, Die Eye Number Determination Apparatus, And Electronic Apparatus Using Same,” TW and US Patent Application Publication, Number: US 7,360,762 B2, Apr.22, 2008.
[3] Ying- Nan Lai, Shun-Tsung Hsu, Chang-Yi Wang, Ming-Tien Tsai, “Method for recognizing dice dots,” US Patent Application Publication, Number: US 2009/0263008 A1, Oct.22, 2009.
[4] Kuo-Yi Huang, “An Auto-Recognizing System for Dice Games Using a Modified Unsupervised Grey Clustering Algorithm,” Sensors 2008, 8(2), 1212-1221.
[5] Lapanja Iztok, Mraz Miha, Zimic Nikolaj, “Computer Vision Based Reliability Control for Electromechanical Dice Gambling Machine,” Industrial Technology 2000, Proceedings of IEEE International Conference on, Vol.2, P.325-328.
[6] 黃國益, 游源洋, 陳成安, 潘志翔, “應用機器視覺於骰子自動計點之辨識與判決系統,” 台灣專利, 專利編號:I272962,中華民國96年2月11日。
[7] Yuval Arie Tresser, Ygael Aron Tresser, “Dice That Recognize The Values Of Their Own Throws And Transmit Them To Computers, With Applications To Electronic And Casino Games,” US Patent Application Publication, Number: US 2005/0215312 A1, Sep. 29, 2005.
[8] 李孟軒, ”骰子點數判別裝置,” 台灣專利, 專利編號: M368475, 中華民國98年11月11日。
[9] 歐陽志宏, “骰子自動辨識裝置,” 台灣專利, 專利編號: 420305, 中華民國90年1月21日。
[10] R. Hartley and A. Zisserman, “Multiple view geometry in computer vision,” Cambridge University, 2nd edition, 2003.
[11] David G. Lowe, “Object Recognition from Local Scale-Invariant Features,” Proceedings IEEE International Conference Computer Vision, vol. 2, 1999, pp.1150-1157.
[12] 吳加山, “使用尺度不變性特徵轉換與立體視覺之即時三維物體識別,” 碩士論文, 國立台灣科技大學, 2008.
[13] Rafael C. Gonzalez, Richard E. Woods, Steven L. Eddins, “Digital Image Processing Using MATLAB,” 1st edition, 2005, 譯者: 繆紹綱.
[14] 鍾國亮, “影像處理與電腦視覺,” 第三版, 2006.
[15] P.N. Belhumeur, J.P. Hespanha, and D.J. Kriegman, "Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection," European Conf. Computer Vision, 1996, pp. 45-58.
[16] Aleix M. MartõÂnez and Avinash C. Kak, “PCA versus LDA,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 23, No. 2, pp. 228-233, February 2001.
[17] S. Baluja H. A. Rowley and T. Kanade, “Neural network-based face detection,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 20, no. 1, pp. 23-38, Jan. 1998.