http://debut.cis.nctu.edu.tw/pages/research/othersC.htm Optical Character Recognition 關於文字辨認（OCR）的研究，本實驗室從成立之初即已開始著手研究，早期在 On-line 方面， 有手寫數學式子的分析辨認與計算，Off-line則有關於手寫郵遞區號的辨認、支票上印 刷體中文字面額的辨認。在手寫中文字的辨認上，也陸續發展出一些筆劃特徵的擷取方法，發 表於學術期刊。 Face Recognition 臉孔辨識在自動化處理的領域中是一門漸漸熱門的學問，隨著資訊世界的發展，電腦的計算 能力已超出人們太多，而相對的，人類的思考，推理，記憶，辨識等能力卻是電腦所望塵莫及的 。因此本實驗室希望能運用影像處理的技術，發展人臉的擷取與辨識系統。 以往的影像處理系統多半是以灰階影像為基礎。在這方面我們有兩位碩士班畢業同學發展了 灰階人臉辨識系統。而近代的影像處理系統多半是彩色高解析度，因此我們期望能以彩色影像為 基礎做人臉的辨識，並且希望藉由兩個攝影機所取得的影像找出人臉立體結構的資訊。 除此之外，我們的臉孔辨識系統必須是容易整合與使用的。因此除了發展人臉辨識的相關演 算法之外，我們也希望能有一些具體的，可應用的展示平台，以供日後讓對臉孔辨識有興趣的人 做參考。 Video Conference Standard 隨著時代的進步，人與人之間遠距離的通訊，由文字變成聲音的傳遞，而在今日，由於通信網 路的發達，以及影像技術的進步，影像的傳送成為新的需求。視訊會議?影像電話也就因應而生了。 近年來由於各種視訊通訊設備不斷出現，由於彼此之間並不相容，造成用戶極大的不便，因此 International Consultive Committee for Telephone and Telegraph (CCITT)便訂定了一國際標準：H.261，做 為開發產品的依據。 本實驗室曾與華邦電子公司合作，在其所生產的W89K、W89K/LC 晶片上發展 了一套合乎H.261標準的編解碼器。 Feature Extraction 本實驗室在早期即有許多傲人成就，在國際學術期刊上，發表多篇論文。發展出許多新的方法 ，利用簡單的數學原理，即可在影像中找出所需要的幾何圖形，如直線、圓形，橢圓形等。 發表論文 OCR L. H. Chen and J. R. Lieh, "Handwritten Character Recognition Using a 2-Layer Random Graph Model by Relaxation Matching" Pattern Recognition, Vol. 23, No. 1, pp.1189-1205, 1990. L. H. Chen, "A New Approach for Handwritten Character Stroke Extraction" Computer Processing of Chinese and Oriental Languages, Vol. 6, No.{LS}1, 1- 17, 1992. L. H. Chen and P. Y. Yin, "A System for On-Line Recognition of Handwritten Mathematical Expressions," Computer Processing of Chinese and Oriental Languages,Vol. 6, No. 1, 19-39, 1992. L. H. Chen and Y. K. Lee, "A New Approach for Recognizing Multifont Chinese Characters Used in a Special Application," International Journal of Pattern Recognition and Artificial Intelligence, Vol. 10, No. 3, 203-222, 1996. L. H. Chen, J. R. Lieh, "Handwritten Character Recognition Using a 2-Layer Random Graph Model by Relaxation Matching," Proceedings of the 1989 Workshop on Computer Vision, Graphics and Image Processing R.O.C., Aug. 1989, pp. 237-248. L. H. Chen and P. Y. Yin, "A System for On-Line Recognition of Handwritten Mathematical Expressions," Proceedings of 1991 International Conference on Computer Processing of Chinese and Oriental Languages, Taipei, Taiwan, 13-15 August, 1991, pp. 138-142. L. H. Chen and Y. K. Lee, "A New Approach for Recognizing Multifont Chinese Characters Used in a Special Application," 11th IAPR International Conference on Pattern Recognition, The Hague, The Netherlands, Aug. 30 - Sept. 3, 1992. Feature Extraction L. H. Chen and W. H. Tsai, "Moment-Preserving Line Detection," Pattern Recognition, Vol. 21, No. 1, pp. 45-53, 1988. L. H. Chen and W.H. Tsai, "Moment-Preserving Curve Detection," IEEE Trans. Syst., Man, Cybern., Vol. SMC-18, No. 1, pp. 148-158, 1988. L. H. Chen and W. H. Tsai, "Moment-Preserving Sharpening: A New Approach to Digital Picture Deblurring," Compt. Vision, Graphics, and Image Process., Vol. 41, No. 1, pp.1-13, Jan. 1988. L. H. Chen and W.H. Tsai, "Moment-Preserving Techniques in Image processing ---A Survey and New Approaches," Proceedings National Science council, ROC(A), Vol. 13, No. 5, pp. 280-301, Sep. 1989. L. H. Chen and S.S. Wang, "A Two-Layer Area-Level Edge Detector--- A New Approach," Journal of Information Science and Engineering, 6, pp. 311-323, 1990. L. H. Chen and K.L. Lee, "A New Method for Circular Object Detection and Location," Pattern Recognition Letters, 11, pp. 691-697, 1990. L. H. Chen and W.H. Tsai, "Survey of Moment-Preserving Techniques in Image Processing," Essays on Computer Vision and Related Topics: In Memory of Dr. K. S. Fu, Julius T. Tou (Eds.), Institute of Information Science Academia Sinica, Taipei, Taiwan, ROC, pp. 388-432, 1990. L. H. Chen, "A Two-Phase Area-Level Line/Edge Detector," Pattern Recognition, Vol. 25, No. 1, 55-63, 1992. L. H. Chen, "A New Approach for Feature Point Classification, Aggregation, and Description," International Journal of Pattern Recognition and Artificial Intelligence, Vol. 6, No. 5 , 849-871, 1992. L. H. Chen and W. H. Tsai, "Location of Curved Edges to Subpixel Accuracy," Journal of the Chinese Institute of Engineers, Vol. 16, No.5, 589-602, 1993. P. Y. Yin and L. H. Chen, "Random-Cumulating Thresholding : A New Approach to Multilevel Thresholding," Signal Processing, Vol. 34, No. 3, 311-322, 1993 C. T. Ho and L. H. Chen, "High Speed Algorithm for Locating Circular Objects," Journal of Electronic Imaging, Vol. 2, No. 2, 108-113, 1993. P. Y. Yin and L. H. Chen, "New Method for ellipse detection by means of symmetry," Journal of Electronic Imaging, Vol. 3, No. 1, 20-29, 1994. C. T. Ho and L. H. Chen, "A Fast Ellipse/Circle Detector Using Geometric Symmetry" , Pattern Recognition, Vol. 28, 117-124, 1995. C. T. Ho and L. H. Chen, "A high speed algorithm for elliptical object Detection", IEEE Transactions on Image Processing, Vol. 5, No. 3, 547-550, March 1996. C. T. Ho and L. H. Chen, "A high-speed algorithm for line detection," Pattern Recognition Letters, 17, 467-473, 1996. L. H. Chen, W. H. Tsai, "Moment-Preserving Line Detection," Proc. Interna. Comput, Symposium, R. O. C., Nov. 1986, pp. 700-709. L. H. Chen, S. S. Wang, "A New Approach for Edge Detection," IEEE International Conference on Image Processing, ICIP'89 Conference Proceedings, Vol. 2, Singapore, Sep. 1989, pp. 631-635. L. H. Chen, S.S. Wang, "A Two-Layer Area-Level Edge Detector--- A New Approach," Proceedings of the 1989 Workshop on Computer Vision, Graphics and Image Processing R. O. C., Aug. 1989, pp. 310-316. L. H. Chen and A. M. Chen, "A High Speed Line/Circle Detector Based on Hough Transform," Proceedings of the Pacific-Rim InternationalConference on Modeling Simulation and Identification, Vancouver, Canada,4-6 August, 1992, pp. 27-30. C. T. Ho and L. H. Chen, "A New Method for Boundary Extraction," IPPR Conference on Computer Vision, Graphics and Imaging Processing, 1992. C. T. Ho and L. H. Chen, "New Method for Circle Detector," Proceedings of the International Conference on Modeling, Simulation and Identification, Wakayama, Japan, 12-16 September, 1994. C. T. Ho and L. H. Chen, "A New High Speed Algorithm for Line Detection." Proceedings of the International Conference on Modeling, Simulation and Identification, Wakayama, Japan, 12-16 September, 1994. ELSE L. H. Chen, K. J. Farn, and C. S. Tsay, "A Distributed Data Allocation Algorithm Based on the Zero-One Goal Programming Model," Journal of the Chinese Institute of Engineers, Vol. 14, No. 5, pp. 551-558, 1991. P. Y. Yin and L. H. Chen, "Random-Cumulating Thresholding : A New Approach to Multilevel Thresholding," Signal Processing, Vol. 34, No. 3, 311-322, 1993 P. Y. Yin and L. H. Chen, "New Method for multilevel thresholding using the symmetry and duality of the histogram," Journal of Electronic Imaging, Vol. 2, No. 4, 337-344, 1993. P. Y. Yin and L. H. Chen, "A new non-iterative clustering approach," Pattern Recognition Letters, Vol. 15, 125-133, 1994. C. H. Lee and L. H. Chen, "A fast motion estimation algorithm based on the block sum pyramid," IEEE Transactions on Image Processing, Nov. 1997. P. Y. Yin and L. H. Chen, "A new iterative scheme for multilevel thresholding," Signal Processing, Vol. 60, Issue 3, 1997 . L. H. Chen, K. J. Farn, and C. S. Tsai, "The Study of Applying the Zero-One Goal Programming to Solve Data Allocation Problem in Distributed System," IX-th International Multicriterion Decision Making (MCDM) Conference, Fairfax, Virginia, 5-8 August 1990.

簡介 What's Steganography? 有關資訊偽裝﹐在學術界上較為常用的術語有 Data Hiding、Information Hiding、Steganography 等三個﹐其中前兩個名詞我們較易從字面上穫得其中的涵意﹐就是將資料或者是資訊隱藏起來的意思﹐兩者如果真要細分﹐可從 Data 與 Information 這兩個字去分辨﹐但在學術界上﹐較注重的是偽裝方法的研 究﹐對於這些名詞間的分別﹐並沒有多去注意。 至於 Steganography 這個名詞﹐則可推溯回西元1499年﹐由 Trithemius 將 steganos 與 graphein 兩個希臘 字合併而成﹐steganos 的字義是 covered﹐是遮蔽、隱藏的涵意﹐graphein 則是 to write﹐兩個字合併起來﹐就是將所寫的字隱藏起來的意思。Steganography 這個字後來因為用法多和魔術、妖術 (Black art) 產生關連﹐所以漸漸較少被用到。直到近來在 David Kahn 所著的 Codebreakers 一書中被賦予新的意義後才又再度起用。 至於密碼學 (Cryptography) 這個字也是由兩個希臘字根所組成﹐Crypto 這個希臘字義是 secret 的意思﹐因此 Steganography 的意義相當接近於 Cryptography﹐也有學者將其視為是 Cryptography 的分支。我們可以用一個簡單的方法來區別這兩個字﹐Cryptography 是將訊息 (Message) 的意義 (Meaning) 隱藏起來﹐讓你看不懂其涵意﹔至於 Steganography 則是將訊息本身的存在性 (Existence) 隱藏起來﹐讓你根本不曉得有訊息的存在。 History of Steganography 史學之父﹐希臘歷史學家 希羅多德 (Herodotus, 484~425BC) 在他的著作曾提到幾個有關資訊隱藏的例子： 第一、 Histiaeus 將奴隸的頭髮剃光﹐然後在頭皮上刺上所要傳遞的秘密訊息﹐ 待其頭髮又長出後才將其送出﹔ 第二、Demaratus 將秘密訊息寫在書寫板的蠟質層下﹔ 第三、Harpagus 將信放置於尚未去皮的野兔中﹐然後將訊息傳遞者喬裝成獵人模樣。 此外，巴達人大約在西元前 450年曾用過一個方法﹐就是在一根桿子繞上一條皮帶﹐然後將訊息順著桿子書寫﹐最後將皮帶解下﹐在空白處隨機填上其它字母。在 Trithemius 的1499年所寫的Steganographia 這本書中也介紹了一種類似的方法﹐就是在一堆無意義的字(word)中﹐如 Parmesiel Oshurmi Delmuson Thafloin Peano Charustrea Melany Lyamunto . 在每兩個字中萃取每兩個字元來﹐透露出以下的拉丁文訊息： Sum tali cautela ut . 還有在十七世紀英國的議會黨與保皇黨之戰 (1642~1652 English Civil War)﹐一位被捕的保皇黨員 John Trevanion 爵士﹐就是靠著從一封被獄卒嚴格檢查的信中﹐在每個標點符號後第二個字母中﹐拼出 Panel at East end of Chapel slides 所以能夠從監牢中脫逃。 在二十世紀﹐Steganography 一直到第二次世界大戰才又受到廣泛的注意﹐美國在所有的郵件與電報通道上過濾所有往來的訊息﹐以防止、偵測間諜的秘密通訊。這些審查員所採用的檢查技術暗示了早期這些隱藏技術設計者(steganographer)所展現的發明才能。 美國稽查局 (The Office of Censorship) 儘可能地禁止所有可能用來傳遞訊息的通道﹐從西洋棋、紡織圖案到小孩子的塗鴨等等都在禁止之列。同樣地﹐縱橫字謎 (Crossword)、剪報 (Newspaper clippings) 也被禁止。郵票因為其面值可以被用來解譯成訊息﹐所以被限制成一共同型式﹔至於空白文件﹐因可能是利用隱形墨水寫成﹐也被禁止。在 1941 年﹐一張僅僅幾平方公釐大小的微點照片，首度被發現在打字機所打出的句點上。一位經過特殊訓練擁有專家知識的稽查人員﹐會用一些相關技術與方法來檢查訊息的真實性﹐檢查後的訊息通常會經過編譯才准以放行﹐用以破壞隱藏於其中的術語 (jargon code)。一個最有名的例子便是在一次世界大戰時﹐一位稽查員將電報內文的 Father is dead. 改成 Father is deceased. 而收信者卻立刻回問：Is Father dead or deceased? 因而露出破綻。 至於報紙﹐則被警告必須小心刊載個人廣告﹐廣播電台的音樂點撥也被限制只能在非特定時段撥出﹐因為商業媒體傳遞秘密訊息的能力﹐已在1940年的一次測試行動被證明有效﹐一位軍事情報軍官在一篇前重量級拳王 Max Baer 的訪問稿中﹐將以下一段訊息隱藏其中 S-112：Queen Elizabeth sails tonight with hundreds of airplanes for Halifax. 並於全國聯播節目中廣播出去。在這次的行動中﹐除了原先就曉得這項秘密的人注意到訊息已被播送出去之外﹐連播音員、電臺經理、Max Baer、甚至全國聯播網上的千萬聽眾都沒能察覺。 DATA HIDING WATERMARKING 發表論文 Conference Papers L. H. Chen and M. S. Liaw "An Effective Data Hiding Method", The conference of CVGIP ,Taiwan. p146 ,1997. 相關連結 N.F.Johnson http://patriot.net/~johnson/Steganography/