图像压缩技术实验程序，图像符号化，256x256图像为例！

VC++编写的可以对图片自由放大或缩小显示的一个程序。-VC can be prepared by the pictures free to zoom in and out a procedure for the show.

各种图像处理方法，包括多种边缘检测算法，也是当初的作业-various image processing methods, including a variety of edge detection algorithm, as well as the original work

几何算法源码(包括多边形填充算法, 多边形裁剪算法和几何绘制算法) 附带演示源码-Geometric algorithm source code (including polygon fill algorithm, polygon clipping algorithm and geometric rendering algorithm) attached demo source

图像处理，图像的检测，模板匹配算法代码，应用与模板搜索-image processing, image detection and template matching algorithm code, application templates and search

GDI编程示例，演示如何建立图形编程环境，如何实现矢量图形的绘制-GDI programme example demonstrate you how to build the setting and how to implement the protract of the vector image.

在开始菜单或桌面上建立快捷方式,方便用户使用快捷地启动软件-the desktop or Start menu to establish efficient, fast and user-friendly software to launch

文件管理系统，对文件进行有效的管理，它将文件以二进制数据的形式存入Access数据库中，也可从数据库中读出文件-Document management system, effective management of the file, it will file in the form of binary data into Access database, can also be read from the database file

一个不错的画图程序-a good drawing program

用directshow进行摄像头图像的捕获和处理。-Camera images using directshow to capture and processing.

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。-Catalog include: documentation and two source code. One of the source code is the identification procedures, and the other is a matrix class library, the book which is also used, so be accompanied. As it has saved the trained network weights, so the first time the use of identification procedures, he can run Release directory of the executable file, and then the picture directory to read the test pictures to identify. Of course, the reader can also be its own training network with training samples, but pay special attention to the training samples, otherwise it may result in the recognition rate is very low. Training sample selection principle is that as much as possible representative, in the long training time Buzhi Yu circumstan