|Table of Contents|

[1] Huang Heming, Da Feipeng, Han Xiaoxu, et al. Wavelet transform and gradient direction based feature extractionmethod for off-line handwritten Tibetan letter recognition [J]. Journal of Southeast University (English Edition), 2014, 30 (1): 27-31. [doi:10.3969/j.issn.1003-7985.2014.01.006]
Copy

Wavelet transform and gradient direction based feature extractionmethod for off-line handwritten Tibetan letter recognition()
基于小波变换和梯度方向的脱机手写藏文字符特征提取方法
Share：

Journal of Southeast University (English Edition)[ISSN:1003-7985/CN:32-1325/N]

Volumn:
30
Issue:
2014 1
Page:
27-31
Research Field:
Computer Science and Engineering
Publishing date:
2014-03-31

Info

Title:
Wavelet transform and gradient direction based feature extractionmethod for off-line handwritten Tibetan letter recognition
基于小波变换和梯度方向的脱机手写藏文字符特征提取方法
Author(s):
Huang Heming1, 2, Da Feipeng1, Han Xiaoxu3
1School of Automation, Southeast University, Nanjing 210096, China
2School of Computer Science, Qinghai Normal University, Xining 810008, China
3Department of Computer and Information Science, Fordham University, New York 10458, USA
黄鹤鸣1, 2, 达飞鹏1, 韩晓旭3
1东南大学自动化学院, 南京210096; 2青海师范大学计算机学院, 西宁 810008; 3福坦莫大学计算机与信息科学系, 纽约 10458
Keywords:
pattern recognition;  wavelet transform;  gradient direction;  Tibetan;  handwritten character
模式识别;  小波变换;  梯度方向;  藏文;  手写字符
PACS:
TP391.4
DOI:
10.3969/j.issn.1003-7985.2014.01.006
Abstract:
To improve the recognition accuracy of off-line handwritten Tibetan characters, the local gradient direction histograms based on the wavelet transform are proposed as the recognition features. First, for a Tibetan character sample image, the first level approximation component of the Haar wavelet transform is calculated. Secondly, the approximation component is partitioned into several equal-sized zones. Finally, the gradient direction histograms of each zone are calculated, and the local direction histograms of the approximation component are considered as the features of the character sample image. The proposed method is tested on the recently developed off-line Tibetan handwritten character sample database. The experimental results demonstrate the effectiveness and efficiency of the proposed feature extraction method. Furthermore, compared with the detail components, the approximation component contributes more to the recognition accuracy.
为了提高脱机手写藏文字符的识别效果, 提出了一种在小波变换基础上计算局部梯度方向直方图的特征提取方法.首先, 对一个脱机手写藏文字符样本图像进行一次Haar小波变换, 得到相应的一级近似分量;然后, 将这个一级近似分量划分成几个等尺寸的子区域;最后, 计算每个等尺寸子区域的局部梯度方向直方图, 并将所有子区域的全部局部梯度方向直方图的值作为该字符图片的特征.在最近建立的脱机手写藏文字符样本数据库(THCDB)上的实验结果表明:提出的特征提取方法识别效率较高, 且识别效果较好;和细节分量相比, 近似分量对提高识别精度具有更大的贡献.

References:

[1] Huang H M, Da F P. General structure based collation of Tibetan syllables [J]. Journal of Computational Information System, 2010, 6(5):1693-1703.
[2] Wang H J, Zhao N Y, Deng G Y. A stroke segment extraction algorithm for Tibetan character recognition [J]. Journal of Chinese Information Processing, 2001, 15(4): 41-46.(in Chinese)
[3] Li Y Z, Wang Y L, Liu Z Z. Study on printed Tibetan character recognition technology [J]. Journal of Nanjing University: Natural Sciences Edition, 2012, 48(1): 55-62.(in Chinese)
[4] Ngodrup, Zhao D C. Research on wooden blocked Tibetan character segmentation based on drop penetration algorithm [C]//Chinese Conference on Pattern Recognition. Chongqing, China, 2010: 84-88.
[5] Liang B, Wang W L, Qian J J. Application of hidden Markov model in on-line recognition of handwritten Tibetan characters [J]. Journal of Microelectronics and Computer, 2009, 26(4): 98-101.
[6] Ma L L, Liu H D, Wu J. MRG-OHTC database for online handwritten Tibetan character recognition [C]//International Conference on Document Analysis and Recognition. Beijing, China, 2011: 207-211.
[7] Huang H M, Da F P. A database for off-line handwritten Tibetan character recognition [J]. Journal of Computational Information System, 2012, 9(18): 5987-5993.
[8] Huang H M, Da F P. Sparse representation-based classification algorithm for optical Tibetan character recognition [J]. Optik-International Journal for Light and Electron Optics, 2014, 125(3):1034-1037.
[9] Huang H M, Da F P. Wavelet and moments based off-line handwritten Tibetan character recognition[J]. Journal of Information and Computational Science, 2013, 10(6): 1855-1859.
[10] Raviraj P, Sanavallah M Y. The modified 2D-Haar wavelet transformation in image compression [J]. Middle-East Journal of Scientific Research, 2007, 2(2): 73-78.
[11] Liu C L. Normalization-cooperated gradient feature extraction for handwritten character recognition [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 29(8):1465-1469.

Memo

Memo:
Biographies: Huang Heming(1969—), male, graduate; Da Feipeng(corresponding author), male, doctor, professor, dafp@seu.edu.cn.
Foundation items: The National Natural Science Foundation of China(No.60963016), the National Social Science Foundation of China(No.17BXW037).
Citation: Huang Heming, Da Feipeng, Han Xiaoxu. Wavelet transform and gradient direction based feature extraction method for off-line handwritten Tibetan letter recognition[J].Journal of Southeast University(English Edition), 2014, 30(1):27-31.[doi:10.3969/j.issn.1003-7985.2014.01.006]
Last Update: 2014-03-20