中文说明：应用背景 BP（Back Propagation）神经网络是1986年由Rumelhart和McCelland为首的科学家小组提出，是一种按误差逆传播算法训练的多层前馈网络，是目前应用最广泛的神经网络模型之一。BP网络能学习和存贮大量的输入-输出模式映射关系，而无需事前揭示描述这种映射关系的数学方程。它的学习规则是使用最速下降法，通过反向传播来不断调整网络的权值和阈值，使网络的误差平方和最小。BP神经网络模型拓扑结构包括输入层（input）、隐层(hidden layer)和输出层(output layer)。 关键技术 通过由数字构成的图像，自动实现几个不同数字的识别，设计识别方 法，有较高的识别率。 作为一个识别系统 , 我们最终要用某些参数来评价其性能的高低 ,
English Description：
Application backgroundBP (BackPropagation) neural network is presented the scientists group 1986 by Rumelhart and McCelland headed, is a kind of error inverse propagation training algorithm for the multilayer feedforward network and is currently the most widely used models of neural network. The BP network can learn and store a lot of input and output mode mapping relationships without prior revealing the mathematical equation describing the mapping relationship.. Its learning rule is the use of the steepest descent method, through the back propagation to adjust the weight and the threshold of the network, so that the error of the network and the smallest. The topological structure of the BP neural network includes the input layer (input) and the hidden layer (hiddenLayer) and output layer (layer output).Key TechnologyThrough the digital image, automatic realization of several different numbers of recognition, design and identification of squa