ACA_FCM

该程序是对原图像进行梯度加掩模的三次分水岭分割(That the procedure was in the original image mask plus three gradient watershed segmentation)

对l1最小化的处理，其中包括全面的l1范数的解得算法，运用tv全变分最小的解决方法，适合于单像素以及图像处理的研究者参考。(L1 minimization, including the full L1 norm solution algorithm, the use of TV total variation, the smallest solution, suitable for single pixel and image processing researchers reference.)

GMSD 基于图像特征的质量评价，matlab编写语言(GMSD image assessment)

各向异性扩散相干斑降噪，可以在增强边缘的同时滤出噪声，效果很好！(Anisotropic diffusion, coherent noise can enhance the edge of the filter out the noise at the same time, the effect is very good!)

对视频生成sprite图像，费了作者很大的力气，拿来与大家共享！(sprite generation of video images, and a lot of effort author, used to share with you!)

大连理工大学卢湖川教授团队显著性检测方法4(Visual Tracking via Probability Continuous Outlier Model)

sift算法在cv领域的重要性不言而喻，该作者的文章引用率在cv界是number1.本篇博客只是本人把sift算法知识点整理了下，以免忘记。本文比较早的一篇博文opencv源码解析之(3)：特征点检查前言1 中有使用opencv自带的sift做了个简单的实验，而这次主要是利用Rob Hess的sift源码来做实验，其实现在的opencv版本中带的sift算法也是Rob Hess的，只是稍微包装了下。 下面来做下试验,试验sift代码采用Rob Hess的代码，opencv目前版本中的sift源码也是采用Rob Hess的。代码可以在他的主页上下载：http://blogs.oregonstate.edu/hess/code/sift/ 这里我下载的是windows版本的，并采用Qt做了个简单的界面。(sift algorithm is self-evident the importance of the cv field, the author of the article referenced in the cv sector is number1 this blog I just sift algorithm knowledge compiled under, so as not to forget. This article compares the earlier blog post opencv source code analysis (3): check the preface of the feature points 1 the opencv bring their own sift to do a simple experiment, but this time Rob Hess' s sift source to experiment. In fact, with the opencv version sift algorithm is Rob Hess, only slightly packing the next. Below to do the next test, test sift code using the code of Rob Hess, sift source opencv version of Rob Hess. The code can be downloaded on his home page: http://blogs.oregonstate.edu/hess/code/sift/ here I downloaded the windows version, and using Qt to do a simple interface.)

说明：   最简单的差分格式有向前、向后和中心3种。 向前差分：f (n)=f(n+1)-f(n) 向后差分：f (n)=f(n)-f(n-1) 中心差分：f (n)=[f(n+1)-f(n-1)]/2(The easiest difference format forward, backward, and three kinds of centers. Forward differencing: f (n) = f (n+ 1)-f (n) Backward difference: f (n) = f (n)-f (n-1) Central difference: f (n) = [f (n+ 1)-f (n-1)]/2)

视频截图的类源码下载

根据Dalal提出的HOG特征算法编写，代码简单，注释清楚，根据所处理图像的名称路径修改后可生成36*105的特征向量。(The algorithm proposed by Dalal HOG feature writing code simple, clear annotation feature vector can generate 36* 105, modified according to the name of the image processing path.)