这份代码经源码格式软件格式化过

一个使用起来十分方便的类，VisObject。对于存取，处理位图很灵活。-to use a very convenient categories, VisObject. For access, processing bitmap very flexible.

本文的目的是借助Paul Viola最新提出的实时特征检测的技术，实现一个能够进行快速人脸检测的系统。并且通过训练，得到尽可能优化的分类器构造参数，从而获得具有高检测速度和检测正确率的人脸检测系统。-The purpose of this paper by drawing on the latest proposed by Paul Viola real-time feature detection technology, to realize a fast face detection system. Through training, to be as much as possible to optimize the classifier structure parameters to obtain a high detection speed and detection rate of correct face detection system.

sclice segmentation algorithm which is the second one programmed by Coo.

基于OpenCV的OCR程序，可以实现数字识别，程序采用KNN算法，可以对样本数据进行计算-Based on OpenCV" s OCR program, you can digital identification process using KNN algorithm can calculate the sample data

一个很不错的图片处理组件，可以进行扫描打印，及相关的图片处理，包含完整原码，如果安装不上的话去下个安装版的，把里的dbk复制过来就可以了。-A very good image processing components, you can scan to print, and related image processing, including integrity of the original code, if not installed, then go on to install the next version, and the inside dbk copied on it.

sobel边缘检测，简单的边缘检测程序，并可进行二值化处理，适合初学者。-edge detection， sobel

Java2DExample.rar Java2D的例子-Java2DExample.rar swing examples

用Java编写的基于Kohonen神经网络的OCR-Java prepared by the Kohonen neural network-based OCR

系统实现方案：总体思想：将图像文件以二进制方式读入缓冲区，把二进制位转制为整数，对缓冲区数据进行编码，以特定的格式(unsigned short)写入输出文件；解压缩过程反过来即可。压缩文件存储格式：以每两个字节为单位储存(即每次写入16位unsigned short类型)。 0   1              2~15                      ... 第0位　存储压缩信息(1压缩；0不压缩，原始数据)第1位　如果第0位为1(压缩)，此位标明字符类别(0或1)；如果第0位为0(原始数据)，此位无意义，默认为0后14位　存储字符数量或者原始数据的值

立体视觉匹配完整源码，可以直接使用，是机器视觉中的重要部分。-three-dimensional visual integrity of the source code matching, can be directly used in machine vision is the important part.