对有噪声的数字 利用神经网络进行识别， 训练效果显著

可以拿来就用的字符识别系统的源代码，识别率比较高啊。-Can be used on the character recognition system used in the source code, the recognition rate is relatively high ah.

此源代码将演示如何用C,Win 32 SDK及GDI+来实现边缘检测。(C++,C)-this source code to demonstrate how to use C, Win 32 SDK GDI and edge detection to achieve. (C, C)

实现盲信号分离的独立成分分析的算法,用Informax实现-the arithmetic realize the ica

这是数字图像处理的一个程序，是将图像均衡化的matlab编程，希望有人喜欢。-This is a digital image processing of a program that is the image equalization matlab programming, want someone to love.

本程序整合了 OSG 大部分的基础功能，程序代码编写采用了模块化思想，有效代码在 1000 行左右，注释比较清晰，对于程序的修改和功能的扩展非常方便。-This procedure combines the foundation for most functional OSG, procedures used to prepare the code modular thinking, effective code line in about 1000, the Notes are more well-defined, for the process to modify and extend the very convenient features.

Only an example of how to print a check pattern for cameras using a borland c++ builder.

VC++NJIN虚拟示波器程序-VC NJIN virtual oscilloscope procedures

clamped bspline function.

vc写的字符识别训练源码，VC编程，解压后有两个压缩包，一个是原代码，一个是可执行文件。-vc written character recognition training source, the VC program, unpacked two compressed, one of the original code, One is the executable file.

提供了图象直方图均衡，robert,sobel,prewitt,kirsch,gauss等方法的边缘检测运算，中值滤波，二维FFT变换和DCT变换，图象的反转，旋转等一些基本的图象模板变换和点变换的功能。目前只处理256*256的图象-the image histogram equalization, robert, segmentation, prewitt, "and Gauss methods of edge detection operators, the median filter, two-dimensional FFT and DCT, image inversion, and rotating some of the basic template image and transform point transformation functions. Currently only handle 256* 256 images