成都大学本科毕业设计（论文）
III
绪 论
...........................................................................................................................................1
1
手写体数字识别的相关技术
.................................................................................................4
1.1 Matlab
平台简述
.......................................................................................................... 4
1.2
人工神经网络
..............................................................................................................4
1.2.1
人工神经网络简述
...........................................................................................4
1.2.2 人工神经网络的基本特征...............................................................................5
1.2.3 人工神经网络的网络模型...............................................................................5
1.3 BP
神经网络
.................................................................................................................5
1.3.1 BP
神经网络简述
..............................................................................................5
1.3.2 BP
神经网络的原理及结构
..............................................................................6
1.3.3 BP
神经网络优缺点
..........................................................................................6
2
手写体数字识别的研究
.........................................................................................................8
2.1 图像预处理...................................................................................................................8
2.1.1 数字图像灰度化处理.......................................................................................9
2.1.2 数字图像平滑滤波...........................................................................................9
2.1.3 数字图像二值化.............................................................................................10
2.1.4 数字图像归一化.............................................................................................10
2.1.5 数字图像细化.................................................................................................10
2.2 手写体数字图像的特征提取....................................................................................11
2.2.1
逐像素特征提取法
.........................................................................................11
2.2.2
骨架特征提取法
.............................................................................................11
2.3
神经网络的搭建
........................................................................................................11
3
手写体数字识别的具体实现
...............................................................................................13
3.1 图像预处理................................................................................................................13
3.1.1 图像灰度化处理.............................................................................................13
3.1.2 图像中值滤波.................................................................................................14
3.1.3
图像二值化处理
.............................................................................................14
3.1.4
图像归一化处理
.............................................................................................15
3.1.5
图像细化处理
.................................................................................................15
3.2
样本图像尺寸的处理
................................................................................................16
3.3
特征值提取
................................................................................................................17
