本程序是基于VisualC++和OpenCV的做的人脸跟踪程序,实现了多人脸跟踪,通过改进Camshift算法,在一定程度上解决了多人脸混淆问题...

直方图修正-彩色变换算法源代码-Histogram amendment- Color Transform algorithm source code

X264(H264)最新版本 包含SVN目录入口,方便更新 注意:下载后请将后缀改名为.tar.bz2 包含win32 makefile 但最好在linux下使用(win32下需修改文档编码为DOS编码)-X264 (H264) the latest version of SVN directory that contains the entrance to facilitate the update Note: Please download suffix renamed. Tar.bz2 contains win32 makefile in linux but it would be best to use (win32 need to modify a document under the code for DOS encoding)

一个简单、快速的图片识别代码，只能识别数字和英文字母。可用于网页上验证码的识别，能识别出验证码常用的各种扰乱的点和线-a simple, quick photo identification code, only an identification number and alphabetical. The website can be used to test code on the identification, certification will identify the various codes used to disrupt the dots and lines

人脸检测的代码，非常有效 人脸检测的代码，非常有效-face detection code, a very effective face detection code, very effective

C++ 语言的图形系统，由风魂设计-C language graphics system, designed by the Wind Spirit

简单的图像处理程序，统计直方图，均值化，用概率法对灰度值进行分类-Simple image processing procedures, statistical histogram, mean, and with the probability of gray value on the classification

可对彩色图像进行行列变化 小波一层 二层 三层变换 还可进行低通 高通 滤波处理 本人亲自试验过了-Color image can change the ranks of two-story three-layer wavelet transform can also be low-pass high-pass filter processing personally tested the

反混淆多边形的绘制程序源码。图形学课上做的实验程序，采用自己设计的算法，而不是用opengl下的反混淆模式-Anti-aliasing polygon drawing program source. Graphics class to do the experimental procedures, the use of their own design algorithm, rather than anti-aliasing under opengl mode

csgl函数库说明文档,包括所有的使用方法-csgl library documentation, including the use of all

为了快速精确地对一个实体进行的三维重建, 提出了一种新的混合建模方法. 该方法利用同一实体在不同视 点下的多幅深度图像, 根据它们之间颜色和纹理的相关性, 通过逆投影变换, 可将图像中象素点映射到三维空间中的正 确位置. 然后再经模型重建和重投影, 可获取实体在任意视点下的图像. 实验结果表明, 这种建模方法具有复杂度恒定、 获取图像真实感强等优点, 适用于形状、 结构复杂的物体的重建.-: In o rder to reconst ruct th ree di mensi on model of an object, a new method of hybr id modeling is p resented . Th is method use mut i p le range i mage of different viewpo int, acco rding to the co lo r and texture, the p ixel po int can be mapped to co r rect po sit i on in th ree di mensi on space by the backp ro ject t ransfo rmat i on. Then the i mage can be go t in any viewpo int by modeling reconst ruct i on and rep ro ject i on. Exper i ments has p roved that th is method can acquire h igh realist ic i mage and has stat ic comp lexit i on. It can be used to reconst ruct some objects w ith comp licated shape and topo logies