论文写作与规范
题目:基于神经网络方法的字符识别方法
学号: 210802102
专业:计算机系统结构
姓名:靳飞飞
2009 年 1 月 9日
基于神经网络方法的字符识别方法
靳飞飞
(中国海洋大学信息科学与工程学院, 山东青岛266071)
摘要:字符识别是模式识别领域的一项传统的课题,这是因为字符识别不是一个孤立的问题,而是模式识别领域中大多数课题都会遇到的基本问题,并且在不同的课题中,由于具体的条件不同,解决的方法也不尽相同,因而字符识别的研究仍具有理论和实践意义。
这里讨论的是用神经网络方法实现基于照相的数字图像的字符识别的问题。
并且通过模板匹配的方法作为参照,以体现神经网络在处理模式识别问题上的优势。
由于人工神经网络的非线性以及并行性和鲁棒性等特点,在上述领域,其取得了以往传统算法无法获得的成功。
关键词:神经网络;字符识别;图像处理
Character recognition based on neural network
Jin Feifei
(College of Information Science and Engineering,Ocean University of China,Qingdao 266071,China)
Abstract:Character recognition is a traditional problem in the field of pattern recognition, for it is rather an isolated task than a fundamental problem in most work of pattern recognition area, with which we have various methods to deal in terms of specific conditions. That means the pursuit of character recognition is of great significance both in theory and in practice .The goal of this paper is using neural network to recognize characters on digital image based on camera. It also can be seen, in the paper, the advantage of neural network compared with the template matching method. Because its nonlinearity, parallel and strong, in these fields mentioned above, artificial neural network has achieved the success which other traditional algorithms can not reach.
Key word: neural network, character recognition, image processing
1引言
字符识别是模式识别领域的一项传统的课题,这是因为字符识别不是一个孤立的问题,
而是模式识别领域中大多数课题都会遇到的基本问题,并且在不同的课题中,由于具体的条件不同,解决的方法也不尽相同,因而字符识别的研究仍具有理论和实践意义。
人工神经网络模式识别方法是近些年提出的新方法,为字符识别研究提供了一种新手段,它具有一些传统技术所没有的优点:良好的容错能力、分类能力强、并行处理能力和自学习能力。
因而,采用神经网络识别方式是一种很好的选择。
参考文献:
[1] Q. Zang, et al. Object Classification and Tracking in Video Surveillance[C]. Canadian Association of Internet Providers, 2008.
[2] S.K. Zhou, et al. Visual Tracking and Recognition using Appearance-Adaptive Models in Particle Filters[J]. IP, 2008, 13(11):1491-1506
[3] C.O. Conaire, et al. Multi-spectral Object Segmentation and Retrieval in Surveillance Video[C]. IEEE International Conference on Image Processing, 2008.
[4] N. Dalal, et al. Human Detection Using Oriented Histograms of Flow and Appearance[C]. European Conference on Computer Vision, 2008.
[5] Y. Freund, et al. A decision-theoretic generalization of on-line learning and an application to boosting[J]. Journal of Computer and System Sciences, 2008, 55(1): 119-139.
[6] J. Friedman, et al. Additive logistic regression: a statistical view of boosting[R]. Dept. of Statistics, Stanford Univ. Technical Report, 2008.
[7] I. Haritaoglu, et al. W4: Real-Time Surveillance of People and Their Activities[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 22(8): 809-830.
[8] A. Lipton. Local Application of Optic Flow to Analyses Rigid versus Non-Rigid Motion[R]. ICCV Workshop on Frame-Rate Vision, 2008.
[9] K. Murphy, et al. Graphical model for recognizing scenes and objects[C]. Neural Information Processing Systems Conference, 2008.
[10] V. Vapnik. Statistical Learning Theory[C]. Wiley Inter-science, New York, 2008.
[11] A. Torralba, et al. Sharing visual features for multi-class and multi-view object detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 29(5): 854-869.
[12] Z. Tu. Probabilistic Boosting-Tree: Learning Discriminative Models for Classification, Recognition, and Clustering[C]. IEEE International Conference on Computer Vision, 2008.。