文件名称:number-recognition
- 所属分类：
- 图形图像处理(光照,映射..)
- 资源属性：
- [Windows] [Visual C] [源码]
- 上传时间：
- 2016-11-05
- 文件大小：
- 171kb
- 下载次数：
- 0次
- 提 供 者：
- 薛**
- 相关连接：
- 无
- 下载说明：
- 别用迅雷下载，失败请重下，重下不扣分！
介绍说明－－下载内容均来自于网络，请自行研究使用
本目录主要包括：文档和两个源代码。
其中一个源代码是识别程序的，另一个是一个矩阵类库的，书里面也使用过，所以一并附上。
因为已经保存了训练好的网络权值，所以第一次使用识别程序时，可以直接运行\\Release目录下的可执行文件，然后对图片目录中的测试图片进行读入、识别。
当然读者也可以自行用训练样本训练网络，不过要特别注意训练样本的选择，否则可能造成识别率很低。训练样本选择的原则是，尽可能的有代表性，在训练时间不至于太长的情况下，训练样本数目尽量多。-This catalog mainly includes: Document and two source code.
One of the source code is to identify the program, and the other is a matrix library, the book which has also been used, so be accompanied.
Because the trained network weights have been saved, the first time you use the recognition program, you can directly run the executable file in the \\ Release directory, and then read and recognize the test pictures in the picture directory.
Of course, the reader can also use the training sample training network, but pay special attention to the choice of training samples, or may result in a very low recognition rate. Training sample selection principle is as representative as possible, in the training time is not too long, the number of training samples as much as possible.
其中一个源代码是识别程序的，另一个是一个矩阵类库的，书里面也使用过，所以一并附上。
因为已经保存了训练好的网络权值，所以第一次使用识别程序时，可以直接运行\\Release目录下的可执行文件，然后对图片目录中的测试图片进行读入、识别。
当然读者也可以自行用训练样本训练网络，不过要特别注意训练样本的选择，否则可能造成识别率很低。训练样本选择的原则是，尽可能的有代表性，在训练时间不至于太长的情况下，训练样本数目尽量多。-This catalog mainly includes: Document and two source code.
One of the source code is to identify the program, and the other is a matrix library, the book which has also been used, so be accompanied.
Because the trained network weights have been saved, the first time you use the recognition program, you can directly run the executable file in the \\ Release directory, and then read and recognize the test pictures in the picture directory.
Of course, the reader can also use the training sample training network, but pay special attention to the choice of training samples, or may result in a very low recognition rate. Training sample selection principle is as representative as possible, in the training time is not too long, the number of training samples as much as possible.
(系统自动生成,下载前可以参看下载内容)
下载文件列表
数字识别系统源代码\Bp.h
..................\ChildView.cpp
..................\ChildView.h
..................\DBpParamater.cpp
..................\DBpParamater.h
..................\DIBAPI.CPP
..................\DIBAPI.H
..................\DigitRec.aps
..................\DigitRec.clw
..................\DigitRec.cpp
..................\DigitRec.dsp
..................\DigitRec.dsw
..................\DigitRec.h
..................\DigitRec.rc
..................\INPUT1.cpp
..................\INPUT1.h
..................\MainFrm.cpp
..................\MainFrm.h
..................\mydiblib.h
..................\ReadMe.txt
..................\..lease\DigitRec.exe
..................\.......\图片\num
..................\.......\....\result.txt
..................\.......\....\whi.dat
..................\.......\....\win.dat
..................\.......\....\测试图片1.bmp
..................\.......\....\测试图片2.bmp
..................\.......\....\测试图片3.bmp
..................\.......\....\测试图片4.bmp
..................\.......\....\测试图片5.bmp
..................\.......\....\测试图片6.bmp
..................\.......\....\测试图片7.bmp
..................\.......\....\训练样本.bmp
..................\res\DigitRec.ico
..................\...\DigitRec.rc2
..................\...\Toolbar.bmp
..................\Resource.h
..................\StdAfx.cpp
..................\StdAfx.h
..................\使用说明.txt
..................\Release\图片
..................\Release
..................\res
数字识别系统源代码