(3.227.208.0) 您好!臺灣時間:2021/04/18 13:46
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:葉日斌
研究生(外文):Yeh, Zin-Bin
論文名稱:中文字識別之前處理研究
論文名稱(外文):Research On Pattern Preprocessing for Chinese Character Recognition
指導教授:林慧珍林慧珍引用關係
指導教授(外文):Hwei-Jen,Lin
學位類別:碩士
校院名稱:淡江大學
系所名稱:資訊工程學系
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:1997
畢業學年度:85
語文別:中文
論文頁數:80
中文關鍵詞:識別前處理識別率線上類神經
外文關鍵詞:recognitionpreprocessingrecognition rateon-lineneural network
相關次數:
  • 被引用被引用:0
  • 點閱點閱:135
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
  時代的進步及科技的發展,促使許多事務皆須依靠電腦來完成.然而
,電腦固然可以幫助人類處理許多事情,卻也相對地為人們帶來了一些新
的困擾.由於電腦是歐美發展的產物,所以其主要的輸入方法是以英文打
字方式的輸入法,這個簡易的輸入法無法直接應用在中文輸入,雖然目前
已有中文鍵盤式的輸入法,然而,對於結構複雜的中文字而言,不論採用
何種輸入法(如倉頡,注音等)都有其缺失與限制性,故為了解決這個問
題,中文自動辨識系統便因應而生.因此如何簡化中文輸入的麻煩,甚至
不須經由鍵盤來輸入,乃是一個很好且重要的研究方向.
有關中文字識別的研究,已有許多方法與成果被發表.然而大部分的中文
文字識別系統中,前處理單元為其重要的模組之一,亦可說是系統的核心
之一,它直接影響到整個識別系統的性能.因此,如何加強前處理單元以
提昇辨識系統的品質,一直是研究者努力的目標.
然而,在中文字辨識過程中所遭遇的問題,大部分是由於個人輸寫習慣的
差異所產生而來的,比如: 文字之不同寫法,筆順變動,即變異相似字等
.而大部分筆順的變動問題是許多線上手寫辨識系統尚未解決的難題,而
本文即以分析字體結構的方式,以筆段為基礎,將中文字內部筆段抽取,
重新排出穩定的次序,做為該文字的標準筆順,如此便可解決一般識別系
統常遭遇的筆順問題.  
本文提出中文字的前處理單元並說明此單元如何修飾某些中文字辨識系統
,並改善其性能及效率.

The relationship between computers and human being is
becoming tighterand tighter. Computers can help people do a lot
of jobs, but also bring some new troubles, one of which is the
input problem of Chinese character. For example,due to the
complicated operating rules, data entry through a keyboard is
notonly expensive but also not capable of inputting documents
with a mixture of text and images. As a result, Chinese
character recognition systems are established for solving this
problem. One of the reasons is the most computersare designed
under the western traditions and philosophy, so comes the
eastern(especially the Chinese character) data entry problem.
Thus, how to simplify the Chinese character input Chinese
character input method becomes a very important research topic.
Handwritten character recognition has enormous scientific and
practicalinterest. Numerous methods have been proposed in the
literature; in most of these methods, the preprocessing unit
plays a very important role and also isone of the recognition
system. It directly influences the recognition rate andquality
of a Chinese recognition system is always the goal of the
research inthe field. Some deviation in written character
affecting the recognition performancemay be due to a variety of
causes, such as noise or different writing habit. However, some
recognition system do not accept a wide variation of patterns.
Forexample, most recognition systems only accept some pre-
confined orderings of strokes for every characters. For this
problem we propose a method of reconstruction of stroke segment
ordering of Chinese characters. It can providea stable order
stroke segment for a written Chinese character, independent of
its written order, to solve above mentioned problems.
Therefore, a specific Chinese character recognition system,
mounted with the proposed preprocessing unit can accept over
free-hand written charactersas well as characters with some
constraints.

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關期刊
 
系統版面圖檔 系統版面圖檔