臺灣博碩士論文加值系統

台灣地區推廣信件郵遞區號分類已有數十年，郵件分揀機器藉由辨識信件郵遞區號可自動化分類郵件。但由於民眾的配合度不夠，仍舊有超過半數的郵件因郵遞區號未書寫在正確位置、未書寫郵遞區號以及未使用標準信封等理由，信件無法使用機器辨識郵遞區號並分類處理。若郵件分揀機器能夠辨識不同大小的信件及地址文字，可大幅提升郵件分揀範圍。本研究目的在於建構具備以上需求之郵件辨識軟體。
郵件辨識可分為郵件影像分析及地址文字辨識二部份。郵件影像分析先以二值化方法處理信件影像，並以連通分量搜尋配合收件地址欄位的位置、長度及寬度條件判斷搜尋信件影像中收件地址欄位所在。
地址文字辨識採用簡單的文字特徵點數及主幹筆劃特徵作為文字辨識的依據，並以地址詞取代傳統逐字辨識方式，藉以提高文字辨識效果及處理速度。

A mail-recognized system is proposed. The proposed system handles with small envelopes and recognizes the destination address. There are two parts in the proposed system: searching the DAB(destination address block) and recognizing the destination address.
In searching the DAB, the source image is binarized and reduced, then the envelope image is searched in the source image. After searching the envelope image, connected component is used to search the DAB in envelope image by position, width and height of connect component.
In recognizing the destination address, feature points and main structure radical of word are used to recognize the address because they are simple and fast. Recognition is matching by a whole city name instead matching two single words of city.

第一章 簡介 1
第二章 文獻回顧 5
2-1 歐美日郵件影像分析相關研究 7
2-2 亞洲地區郵件影像分析相關研究 9
2-3 文字分割 13
2-4 中文字識別 18
2-5 結論 23
第三章 系統規劃 26
3-1 郵件影像分析軟體與郵件分撿機 26
3-2 郵件影像分析 30
3-3 地址文字辨識及後處理 39
第四章 地址資料庫設定 47
4-1 樣本收集 47
4-2 特徵點數及筆劃數地址文字資料庫建立 49
4-3 主幹筆劃特徵資料庫建立 60
4-4 地址辨識方法設定 77
第五章 實驗結果與討論 79
5-1 郵件信封的格式以及使用設備規格 79
5-2 辨識結果 82
第六章 結論與建議 94
參考資料 96
附錄A 100
附錄B 105

Cheng, F. H., “Multi-stroke Relaxation Matching Method for Handwritten Chinese Character Recognition”, Pattern Recognition, vol.31 no.4, pp.401-410, 1998.
Cheng, F. H., “Point pattern matching algorithm invariant to geometrical transformation and distortion”, Pattern Recognition Letters, vol.17, pp.1429-1435.
Downton, A. C. and Leedham, C. G., “Preprocessing and presorting of envelope image for automatic sorting using OCR”, Pattern Recognition, Vol.23, No.3/4, pp. 347-362, 1990.
Gonzalez, R. C. and Woods, R. E., “Digital Imaging Process”, Addison-Wesley, 1992.
Hilderbrandt, T. H. and Liu, W., “Optical recognition of handwritten Chinese characters: advances since 1980”, Pattern Recognition, vol.26, no2. pp205-225, 1993.
Jain, A. K. and Bhattacharjee, S. K., “Address block location on envelopes using Gabor filters“, Pattern Recognition, vol.25, no.12, pp.1459-1477, 1992.
Lee, S. W. and Kim, K. C., ”Address block location on handwritten korean envelopes by the merging and splitting method”, Pattern Recognition, vol.27, no.12 pp.1641-1651, 1994.
Palumbo, P. W. and Srihari, S. N., “Postal address reading in real time”, International Journal of Imaging Systems and Technology, vol.7, pp.370-378, 1996.
Seni, G. and Cohen, E., “External word segmentation of off-line handwritten text lines”, Pattern Recognition, vol.27, no.1, pp.41-52, 1994.
Sossa, J. H., “An improved parallel algorithm for thinning digital patterns”, Pattern Recognition Letters, vol.10, pp.77-80, 1989.
Srihari S. N., etc, “Recognizing address blocks on mail pieces”, Artificial Intelligence Magazine, vol.12, pp.25-40, 1987.
Su, Y. M. and Wang, J. F., “Recognition of handwritten Chinese postal address using neural networks”, IEEE Circuits and Systems, 1998.
Tseng, L. Y. and Chuang, C. T., “An efficient knowledge-based stroke extraction method for multi-font Chinese characters”, Pattern Recognition, Vol.25 No.12 pp.1445-1458, 1992.
Tseng, L. Y. and Chen, R. C., “Segmenting handwritten Chinese characters based on heuristic merging of stroke bounding boxes and dynamic programming”, Pattern Recognition Letters, vol.19, pp.963-973, 1998.
Wang, A. B. and Fan, K. C., “Optical recognition of handwritten Chinese characters by hierarchical radical matching method”, Pattern Recognition, vol.34 no.1 pp.15-35, 2001.
Xue, J. etc, “Destination address block location on handwritten Chinese envelopes”, IEEE Document Analysis and Recognition, 2000.
Xue, J. etc, “Location and interpretation of destination addresses on handwritten Chinese envelopes”, Pattern Recognition Letters, vol.22 pp.639-656, 2001.
Yeh, P. S., Antoy, S., Litcher, A. and Rosenfeld , A., “Address location on envelopes”, Pattern Recognition, vol.20, no.2, pp.213-227, 1987.
Zhang, T. Y. and Suen, C. Y., “A fast parallel algorithm for thinning digital patterns”, Communication of ACM, vol.27, no.3, pp.236-239, 1984.
交通部郵政總局，「台灣地區郵遞區號簿」，交通部郵政總局印製，台北，(1994)。
呂志仁，「郵件分揀機多專家郵遞區號辨識系統」，國立成功大學資訊工程研究所碩士論文，1993。
施紹謙，「利用影像分析技巧對機械製圖作自動瞭解」，國立交通大學資訊科學研究所碩士論文，1988
黃百綱，「中文郵件地址辨識」，國立交通大學資訊科學研究所碩士論文，1994。
郭俊谷，「以線段合併為基礎進行機械製圖辨識之研究」，國立清華大學動力機械工程學系研究所碩士論文，1998