i---retinex

这是一重要的图像修补源代码，基于图像快的修补，是修补算法的经典制作（1）(inage quilting image inpainting )

说明：  利用帧间差分法对视频中的目标进行检测，动态，可运行(Target detection in video)

平常经常用到的一个小程序，两个矩阵相乘的求法(Common often USES a small program, two of the matrix multiplication is also given)

自动曝光的一种算法，该算法程序简单，效果好。令附有效果图(Automatic exposure an algorithm, which is a simple procedure, with good results. Order accompanied by renderings)

运用最小二乘和一范数约束求解低分辨率多光谱图像和高分辨率全色图像融合成高分辨率多光谱图像，程序附带主流的图像融合评价标准，需要融合的图像，对比图像；以及附带程序所需的稀疏包(Using least squares and a norm constraint solving low-resolution multi-spectral image and panchromatic image fusion into a high-resolution multi-spectral images, the program comes with mainstream image fusion uation criteria require the integration of image contrast image and Sparse package comes with the program required)

快速剪切波变换的源程序，用于图像处理，运行速度快(Fast shear wave transform source, used for image processing, fast running)

说明：  机器视觉系统实例--颜色识别系统，matlab实现(Examples of machine vision systems- the color identification system, matlab achieve)

设计了完整的基于颜色直方图的图像检索系统，效果还算可以(Design of a complete color histogram-based image retrieval systems, the effect can be fairly)

gabor特征提取程序，对于搞问纹理提取的人来说很有用处(gabor feature extraction process, engage in asking for the texture extraction is very useful for those who)

cvpr2012_oral On Multiple Foreground Cosegmentation(In this paper, we address a challenging image segmentation problem called multiple foreground cosegmentation (MFC), which concerns a realistic scenario in general Webuser photo sets where a finite number of K foregrounds of interest repeatedly occur over the entire photo set, but only an unknown subset of them is presented in each image. This contrasts the classical cosegmentation problem dealt with by most existing algorithms, which assume a much simpler but less realistic setting where the same set of foregrounds recurs in every image. We propose a novel optimization method for MFC, which makes no assumption on foreground configurations and does not suffer from the aforementioned limitation, while still leverages all the benefits of having co-occurring or (partially) recurring contents across images. Our method builds on an iterative scheme that alternates between a foreground modeling module and a region assignment module, both highly efficient and scalable. In part)