摘要
人工识别阿拉伯数字既耗时又费力,提出了一种利用BP网络与竞争网络相结合的神经网络来识别阿拉伯数字的模型,仿真结果表明,该模型能够对数字进行准确、快速地识别,且抗噪能力较好.可以用于对部队番号、邮政编码、证件等进行识别.
Artificially identifying Arabic numbers very wastes time, and therefore a model for intelligently identifying Arabic numbers by combining BP neural networks with competitive network is put forward. The simulation result shows that this model can quickly and correctly identify Arabic numbers, and it has better anti-noise ability. It can be used for identifying the code designation of a military, postal code and credentials code.
出处
《西安石油大学学报（自然科学版）》
CAS
2006年第2期69-72,共4页
Journal of Xi’an Shiyou University（Natural Science Edition）
基金
空军工程大学学术基金资助项目(编号:2004AA780101-03)
关键词
人工神经网络
数字识别
BP算法
噪声
artificial neural network
identification of Arabic numbers
noise
