数字识别系统源代码:

第一步：训练网络。使用训练样本进行训练

第二步：识别。首先，打开图像（256色）；再次，进行归一化处理，点击“一次性处理”；最后，点击“R”或者使用菜单找到相应项来进行识别。识别的结果显示在屏幕上，同时也输出到文件result.txt中。



该系统的识别率一般情况下为90％。



此外，也可以单独对打开的图片一步一步进行图像预处理工作，但要注意，每一步工作只能执行一遍，而且要按顺序执行。

具体步骤为：“256色位图转为灰度图”－“灰度图二值化”－“去噪”－“倾斜校正”－“分割”－“标准化尺寸”－“紧缩重排”。-digital identification system source code : the first step : training network. Training samples used for training the second step : identification. First, open the Image (256 colors); Again, a naturalization processing, click on "one-time"; Finally, Click on the "R" or use the menu item corresponding to find identification. The results show identification on the screen, but also the output to a file which result.txt. The system's recognition rate under normal circumstances 90%. Alternatively, you could open a separate pair of step-by-step photographs for image preprocessing work, but bearing in mind that each step can only be implemented again, but according to the implementation of the order. Specific steps : "256-color bitmap to grayscale" - "graysc