Minnimum-error-thresholding

基于小波变换的人脸识别代码本代码对图像进行小波分解，然后又用最近邻方法实现了图像的识别。(Face Recognition Based on Wavelet Transform code-wavelet based face recognition, the code of the image wavelet decomposition, and then with the nearest neighbor method to achieve the image recognition!)

说明：  基于遗传算法的图像融合，应用于求解图像融合的常用算法，matlab编写 ，很好的实现了，希望 对大家学习 有用(GA,Image Fusion Based on Genetic Algorithms)

比较反滤波和维纳滤波对一幅图像复原的效果。从两方面出发，一是对运动模糊和均值模糊复原的能力，二是抗噪能力，各种情况分别都进行了对比(Comparison of inverse filtering and Wiener filtering on an image restoration results. Proceed in two ways, first on the motion blur and the mean fuzzy recovery capability and noise immunity, all cases are compared separately)

介绍了不同非线性模型在随机共振中的应用，像饱和非线性，阈值非线性等(Describes the different nonlinear model of stochastic resonance in the application, such as nonlinear saturation threshold nonlinearity. . . .)

医学图像处理方面的一些基础算法，VC环境下编译的一个非常好的程序框架(Medical image processing, some of the basic algorithm, VC compiled under the framework of a very good program)

分别利用 Roberts、Prewitt和 Sobel 算子对原图像进行边缘检测和锐化。(Respectively use Roberts, Prewitt and Sobel operator on edge detection and sharpening the original image.)

说明：  针对现有方法存在的隐藏容量低、操作不灵活的缺陷，文章提出了一种基于无损压缩的加密图像可逆 信息隐藏算法。(In view of the defects of low hiding capacity and inflexible operation of the existing methods, this paper proposes a reversible information hiding algorithm for encrypted images based on lossless compression.)

说明：  图像处理,matlab文件，利用SURF进行图像拼接，完成图像配准(Image processing, matlab files, image stitching using SURF, image registration completed)

本程序是用似然比检验的方法进行profile中的变点检测(This procedure is used likelihood ratio test method of change point detection profile)

说明：  MATLAB如何将图片切分成任意M*N块 在图像处理的时候，有时会不可避免的将图像分块来进行操作，但是图像的像素不可能每次都是可以被整除的，所以我们需要imresize函数对图像的尺寸进行修改，以便可以完成M*N块的分割(How can matlab cut the image into any m * n blocks? Sometimes it is inevitable to divide the image into blocks for operation. However, the pixels of the image can not be divided every time. So we need to modify the size of the image with the function of reduce to complete the m * n block segmentation)