Font Size: a A A

Classiifcation And Recognition Technology Research In Print Uighur Recognition System

Posted on:2013-02-28Degree:MasterType:Thesis
Country:ChinaCandidate:Q ChenFull Text:PDF
GTID:2218330374466457Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
The character recognition is an important application of pattern recognition. Atpresent, the unique writing the Arabic script and Arabic script-based Uighur text, aswell as its own structure, complexity and characteristics of the Uighur text recognitiontechnology is relatively small and not very mature. Development of Uighur characterrecognition technology plays an important role in the study of minority history andculture of China's western region, religion, ancient literature and texts, thedevelopment of Arabic text recognition technology also has some reference value. Inorder to accelerate the construction of the information, the deep research in this areaand work is necessary.By analyzing the writing characteristics and structural features of Uighur and thedifficulties of its identification, we studied the Uighur text classification andrecognition technology exploratory and use the VC++to achieved it. The works thatwere done in this article are mainly the following aspects:1. Describing the background and significance of the subject characteristics andpre-technical part of the Uighur gives a brief introduction;2. Selecting characteristics, including the ratio of foreground and background pixels,the form of alphabetic writing, letters, number of strokes, the ring structure as well asthe baseline position single letters image in is described in detail, and the coding ofthese features extraction, so as to provide a basis for classification and identificationof the Uighur alphabet;3. Designing and realizing of the discriminant rules based four classifier, theclassifier Uighur alphabet is divided into52classifications, and using templatematching recognition method to identify the dimensional alphabet, the algorithm andits implementation process are described in detail; The experimental results show that the recognizing methods for recognizing theUighur is effective, the correct recognition rate of this system is94%.
Keywords/Search Tags:Printed Uyghur, Recognition, Feature extraction, Template matching
PDF Full Text Request
Related items