pingjia

28个vc++图像处理原程序之--对比度拉伸(28 vc++ Image processing of the original procedure- contrast stretching)

光学车间加工，很好很强大的一本书。里面有不少值得学习的东西。(Optical shop testing)

灰度双线性插值，缩放图像。可以对图像进行几乎无损的小倍数放大，倍数大时有损。(Gray bilinear interpolation, zooming the image. Can be almost non-destructive image to enlarge small multiples, multiples of large detrimental.)

图像库中有240幅完整的和残缺的飞机和舰船遥感图像。可用hu矩对其特征提取，用于分类。(Image library has 240 complete and incomplete aircraft and ships remote sensing images. Available hu moments of its feature extraction for classification.)

说明：  VC++功能较多的图像处理程序，实现的功能有：图像变换 图像配准 图像分割、图像编码、图像增强、图像复原、图像特征码提娶图像识别、运动检测、图像显示等，特别是源码里面的一些VC的图像处理类，你可以用到你自己的程序中。(VC ++ image processing program with more functions. The functions implemented are: image transformation, image registration, image segmentation, image encoding, image enhancement, image restoration, image feature code, image recognition, motion detection, image display, etc. Some of the VC image processing classes, you can use it in your own programs.)

基于随机Hough变换的圆检测方法，可检测不同位置、半径的圆。(Based on random Hough transform circle detection method can detect a different location, the radius of the circle.)

基于遗传模拟退火和Otsu法的图像分割。采用一种基于模拟退火算法(SA)的混合遗传算法(GASA),来进行Otsu方法选择阈值，其中进行求解时是非线性的遗传算法（GA）的应用优化了求阈值过程，并尽可能地减少其运算量,而加入SA则避免了遗传算法(GA)常见的早熟收敛现象,从而实现有效简单的图像分割。 (adopt hybrid genetic algorithm(GASA) based on simulated annealing algorithm,choose threshold by Ostu method,which applicated the non-linear genetic algorithm(GA) allows us to optimize the process of threshold. And, where possible, to reduce the computation. in order to achieve effectivly and simply image segmentation. Combination SA is to avoid the common phenomenon about premature convergence in genetic algorithm (GA) to achieve a simple and effective image segmentation.)

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)

一种基于non-sample的纹理合成demo，程序性能还有待优化，效果还不错.(texture synthesis demo.)

此代码应用于图像分割，是基于类间方差的大津法求分割阈值(This code applies to image segmentation is based on the variance between two classes of Otsu threshold segmentation uated)