基于特征行必要-充分性匹配的字符识别方法

被引：6

作者：

李佐

王姝华

蔡士杰

机构：

[1] 南京大学计算机软件新技术国家重点实验室

[2] 南京大学计算机软件新技术国家重点实验室江苏南京

来源：

软件学报 | 2002年 / 01期

关键词：

计算机图形学; 字符识别; 特征抽取; 特征行; 必要-充分性匹配;

D O I：

10.13328/j.cnki.jos.2002.01.012

中图分类号：

TP391.4 [模式识别与装置];

学科分类号：

0811 ; 081101 ; 081104 ; 1405 ;

摘要：

字符识别系统的性能在很大程度上依赖于所选取的字符特征.提出了一种基于特征行必要-充分性匹配的OCR(optical character recognition)方法.该方法使用字符模板的特征行集,通过对待识字符位图进行必要性和充分性双向匹配来识别字符.并采用基线对齐归一化方法在特征匹配时准确定位,使识别率和识别速度都较为理想.同时,对字符位图特征行的交互选择、测试和调整等方法做了详细介绍.另外,还提出了字符骨架与位图重叠显示的方案,有效地提高了对识别结果校对的速度.最后,通过测试和比较,对识别效率进行了分析.

引用

页码：85 / 91

页数：7

共 8 条

[1]

Feature extraction methods for character recognition-a survey. Trier OD,Jain AK,Taxt T. Pattern Recognition . 1996

[2]

A document image understanding system for tender. Wang,Shu-hua,Li,Zuo,Yang,Ruo-yu, et al. Proceedings of theInternationalSymposium onFutureSoftwareTechnology? . 1999

[3]

A Feature for Character Recognition Based on Directional Distance Distributions. Ilseok O,,Suen C Y. Proceedings of the 4th IEEE International Conference on Document Analysis and Recognition . 1997

[4]

Analysis of class separation and combination of class-dependent features for handwriting recognition. Oh,II-Seok,Lee,Jin-Seon,Suen,ChingY. IEEE Transactions onPatternAnalysis andMachineIntelligence . 1999

[5]

Decision Combination in multiple classifier systems. T.Ho,J.Hull,and S.Srihari. IEEE Transactions on Pattern Analysis and Machine Intelligence . 1994

[6]

A heuristic algorithm for the recognition of printedChinese characters. Chuang,Chen-Tsun,Tseng,Lin-Yu. IEEE Transactions onSystems,Man, andCybernetics . 1995

[7]

Recognition of unstrained handwritten numerals by double self-organizing neural network. CHO S B. Proceedings of the 13th International Conference on Pattern Recognition . 1996

[8]

Nonuniformly Sampled Feature Extraction Method for Kanji Character Recognition. K.Yamada. Proceeding of the International Conference on Document. Analysis and Recognition . 1997

← 1 →