gh

图像增强的算法，但是有一定的局限性……但是对于一般比较典型的问题效果还是不错的……(Image enhancement algorithms, but there are some limitations ... but for a typical effect is good ...)

程序设计步骤： 1、去均值 2、计算协方差矩阵及其特征值和特征向量 3、计算协方差矩阵的特征值大于阈值的个数 4、降序排列特征值 5、去掉较小的特征值 6、去掉较大的特征值(一般没有这一步） 7、合并选择的特征值 8、选择相应的特征值和特征向量 9、计算白化矩阵 10、提取主分量(Programming steps: 1 to mean 2, calculation of covariance matrix and its eigenvalues and eigenvectors 3, the calculation of covariance matrix eigenvalues greater than the threshold number of 4, in descending order of eigenvalue 5, remove the smaller eigenvalue 6, remove the larger eigenvalue (usually not this step) 7, the combined choice of eigenvalues 8, select the appropriate eigenvalues and eigenvectors 9, computing whitening matrix 10, principal component extraction)

下面给出了48项泽尼克多项式，外加一项常数项。需要注意的是，读者并不需要严格按照下文所示的顺序排列这些泽尼克项，实际上在不同的应用和机构会采用不同的排列顺序。 表中的#0项是个常数或者说是平移项（piston term），这一项的系数也代表了平均光程差；而#1和#2项分别是x和y方向的倾斜项（tilt terms），#3代表了聚焦，因此，#1到#3项代表了波前的高斯或者近轴特性；#4和#5项代表了像散和离焦，#6和#7项代表彗差和倾斜，而#8项代表了3级像差和离焦，也就是说#4到#8项为3级相差项；同样地，#9到#15项代表了5级像差，而#16到#24项代表了7级像差，#25到#35项代表了9级像差，#36到#48项代表了11级像差。(Here are 48 Zernike polynomials, plus a constant term. It is important to note that the reader does not need to strictly arrange the Zernike entries in the order shown below, but in fact different orders of arrangement are applied to different applications and institutions. Table #0 is a constant or a shift (piston term), a coefficient of this term also represents the average optical path difference; while #1 and #2 are inclined X and Y direction (tilt terms), #3 represents the focus, therefore, #1 to #3 on behalf of the wavefront. Gauss or paraxial properties; #4 and #5 represent the astigmatism and defocus, #6 and #7 represent coma and tilt, and #8 represents the 3 order aberration and defocus, that is to say #4 to #8 is 3 level difference; similarly, #9 to #15 on behalf of the 5 order aberration #16 and #24, to represent the 7 level #25 to #35 aberration, on behalf of the 9 #36 to #48 aberration, on behalf of the 11 order aberration.)

说明：  该代码为Harris角点检测，两帧图像匹配的matlab代码(The code for Harris corner detection, two images matching matlab code)

说明：  相位恢复算法,gs算法，适用于双强度测量的相位恢复问题(a phase retrieval algorithmn-gs using MATLAB)

用于姿态初始化（position initialize）(for position initializing)

Graph Cutshi是最近在计算机视觉中研究和应用比较多的优化算法，典型的用于stereo matching, image restortion, texture synthesis等应用，以上代码给出了graph cut，mean shift, 以及相关的ppm/pgm图像借口，本来是用于stereo matching的，已经把相关的matching的实现代码给去掉了。大家可以以此为平台做一些相关的研究和应用。 (Graph Cutshi recently in computer vision research and application of a relatively large number of optimization algorithms, typical for the stereo matching, image restortion, texture synthesis applications, the above code gives the graph cut, mean shift, and the associated ppm/pgm images an excuse, it is used for stereo matching, and has been related to matching the code to realize removed. U.S. can also be used as a platform to do some related research and applications.)

基于小波包图像去噪 很好用的程序 小波去噪(Good use of the procedure of wavelet denoising based on wavelet packet image denoising)

图像的三种分水岭分割方法，一般分水岭分割，使用梯度的两次分水岭分割及使用梯度加掩模的三次分水岭算法(images of the three watershed segmentation method, the general watershed segmentation, gradient of the two watershed segmentation and the use of the gradient increases mask the three watershed algorithm)

说明：  去噪模型TV,可用于图像去噪，效果良好，模型特别简单(The denoising model TV can be used for image denoising, the effect is good, and the model is particularly simple.)