首页> 外文期刊>Pattern Recognition: The Journal of the Pattern Recognition Society >Contour-based handwritten numeral recognition using multiwavelets and neural networks
【24h】

Contour-based handwritten numeral recognition using multiwavelets and neural networks

机译:使用多小波和神经网络的基于轮廓的手写数字识别

获取原文
获取原文并翻译 | 示例
           

摘要

In this paper, we develop a handwritten numeral recognition descriptor using multiwavelets and neural networks. We first trace the contour of the numeral, then normalize and resample the contour so that it is translation- and scale-invariant. We then perform multiwavelet orthonormal shell expansion on the contour to get several resolution levels and the average. Finally, we use the shell coefficients as features to input into a feed-forward neural network to recognize the handwritten numerals. The main advantage of the orthonormal shell decomposition is that it decomposes a signal into multiresolution levels, but without down-sampling. Wavelet transforms with down-sampling can give very different coefficients when the input signal is shifted. This is the main limitation of wavelet transforms in pattern recognition. For the shell expansion, we prefer multiwavelets to scalar wavelets because we have two coordinates x and y for each point on the contour. If we extract features from x and y separately, just as Wunsch et al. did (Pattern Recognition 28 (1995) 1237), then we may not get the best features. In addition, we know that multiwavelets have advantages over scalar wavelets, such as short support, orthogonality, symmetry and higher order of vanishing moments. These properties allow multiwavelets to outperform scalar wavelets in some applications, e.g. signal denoising (IEEE Trans. Signal Process. 46 (12) (1998) 3414). We conducted experiments and found that it is feasible to use multiwavelet features in handwritten numeral recognition. (C) 2003 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved. [References: 18]
机译:在本文中,我们使用多小波和神经网络开发了手写数字识别描述符。我们首先跟踪数字的轮廓,然后对轮廓进行归一化和重新采样,以使其不变且缩放不变。然后,我们在轮廓上执行多小波正交壳展开,以获得几个分辨率级别和平均值。最后,我们使用壳系数作为特征,输入到前馈神经网络以识别手写数字。正交壳分解的主要优点是,它可以将信号分解为多分辨率级别,但无需下采样。当输入信号发生移位时,具有下采样的小波变换可以给出非常不同的系数。这是小波变换在模式识别中的主要局限性。对于壳展开,相对于标量小波,我们更喜欢多小波,因为轮廓上的每个点都有两个坐标x和y。如果我们分别从x和y提取特征,就像Wunsch等人一样。 (Pattern Recognition 28(1995)1237),那么我们可能无法获得最佳功能。另外,我们知道多小波比标量小波具有优势,例如支持短,正交性,对称性和消失矩更高阶。这些特性使多小波在某些应用中的性能优于标量小波。信号降噪(IEEE Trans。Signal Process。46(12)(1998)3414)。我们进行了实验,发现在手写数字识别中使用多小波特征是可行的。 (C)2003模式识别学会。由Elsevier Science Ltd.出版。保留所有权利。 [参考:18]

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号