配准校正

随机森林用于 图像处理matalab 代码(Random Forest matalab code for image processing)

图像匹配最新的源码 matlab 适合初学者和进阶者作者hongshengli (The latest source of image matching matlab for beginners and advanced by Author hongshengli)

聚束SAR极坐标格式算法，包括dechirp处理以及距离向插值和方位向插值(Spotlight SAR polar format algorithm, including dechirp processing and distance and azimuth to the interpolation Interpolation)

本案例采用基于块匹配的图像拼接流程来执行拼接操作，载入图片文件夹来作为待拼接对象，通过进行图片序列的匹配、融合来得到拼接效果，并分别对灰度图像、彩色图像进行处理。(This case is based on block matching process to perform image stitching splicing operation, load images folder as an object to be spliced, by matching the fusion image sequence to get the mosaic effect, respectively, on the gray image, color image processing .)

调用 envi5.3自带的FMask云检测功能，并输出掩膜文件(Call the FMash cloud detection function of envi5.3 and output the mask file)

用python实现图像的LBP特征提取，参数可以自己修改。(Get a LBP image original binary image with python.)

发表于ECCV上的一篇用于人脸识别的算法，Gabor Feature based Sparse Representation for Face Recognition with Gabor Occlusion Dictionary (At ECCV on an algorithm for face recognition, Gabor Feature based Sparse Representation for Face Recognition with Gabor Occlusion Dictionary)

应用到了操作系统界面功能支持库和图像格式转换支持库，在截取窗口位置时调用了API函数，大家做屏幕截图时可以参考下本例程中的代码。(Applied to the operating system interface features support libraries and image format conversion support library , in the interception of the window position API function calls , you can refer to the code in this routine when we do screenshots .)

使用MATLAB生成二维voronoi图的代码，可以用作后续二维voronoi的多晶模型建立(The code of two-dimensional Voronoi diagram generated by MATLAB can be used to build the polycrystalline model of two-dimensional voronoi.)

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)