CRF2D

利用匹配滤波器实现图形识别，包括程序和说明文档(Graphic Recognition Using Matched Filter)

彩条产生程序。。。。720p需添加74.25M时钟(colorbar generation. need 74.25mhz clock if 720p gen)

SVM实现数据分类，可自行选取核函数，用于图像分割操作。程序可自动训练分类平面。(SVM for data classification, are free to select the kernel function，For image segmentation operation. The program can automatically train classifiers plane.)

Localization of inner and outer boundary of iris(Localization of inner and outer boundary of iris,)

说明：  自己编写的图像质量评价函数，包括方差、平均梯度、模糊熵、信息熵(I have written the image quality evaluation function, including the variance, with an average gradient, fuzzy entropy, information entropy)

图像的几何变换，基于VC++的MFC的。旋转、放大、缩小等(Image geometric transformation, based on the VC++ the MFC. Rotate, zoom in, zoom out, etc.)

利用灰度积分投影直接对人脸图像进行检测和眼睛定位是一种常用的算法,但是直接采用该算法会受到背 景、特征等因素的影响,识别准确率较低。提出了一种基于最大类间方差阈值和区域膨胀相结合的检测与定位算法。该算 法首先计算最大类间方差设置阈值,把灰度图像转换为二值图像并检测出人脸区域,然后通过对该人脸区域中的连通区域 进行膨胀及连通性处理,精确定位眼睛坐标。实验表明,此算法可靠,具有较好的识别效果。(face recognized)

这是sift特征提取源码，详细介绍了算法原理及应用，对图像图形处理专业人员的重要参考价值！(This is the sift feature extraction source, and detailed algorithm and application of important reference value for image and graphics professionals!)

一个外国人写的基于分割的立体匹配算法代码，可以在matlab7.1以后版本中直接执行，代码结构也很清晰，适合分析学习(Written by a foreigner based on the partition of the three-dimensional matching algorithm code, you can later matlab7.1 direct the implementation of the code structure is very clear and suitable for analysis of learning)

这是一段视频分割的源代码 效果比较好 希望对大家有帮助(This is a video segmentation of the source code for the effect we hope to help)