LMMSE

说明：  基于radon正反变换的抛物线型多次波去除(Removal of parabolic multiple waves based on Radon transform)

vc获取指定文件的图标的程序源码，希望对大家有所帮助(vc for the specified source file' s icon in the program, we hope to help)

图像识别，包括人脸识别和虹膜识别，在数字图像中虹膜位置的有效定位是虹膜识别的关键问题。用一种基于主动轮廓线模型的方法定位虹膜的位置，先用灰度投影法检测出瞳孔内的一点作为瞳孔的伪圆心，该圆心只要能落在瞳孔内部即可。(Image recognition, including face recognition and iris recognition, iris position in the digital image of the iris recognition and effective positioning is the key issue. Using a model based on active contour method of positioning the iris position, the first gray projection method used to detect a point within the pupil as a pupil of the pseudo-center of a circle, the center of a circle can be as long as it falls within the pupil.)

这时高斯模糊的源码，对做图像盲复原的同学可以参考(Then Gaussian blur of the source, to make the image blind restoration of the students can refer to the)

说明：  用OSTU和KSW实现多阈值图像分割，matlab代码(Image Segmentation Using OSTU and KSW)

压缩感知compressed sensing ,剑桥大学 教授，可以用于降噪、图像处理、文件压缩等等；(Cambridge- Compressive Sensing- Theory and Application；)

基于三维递归搜索的运动估计得到运动矢量后进行运动补偿帧插算法，最后得到视频帧速率上转换序列(frame rate up conversion based on motion estimation)

在指纹特征提取中，结合方向图和图像局部灰度方差对指纹图像进行分割，然后再进行阈值分割去处边缘效应得出较好的分割结果。(In the fingerprint feature extraction, the combination of pattern and image local variance of gray-scale fingerprint image segmentation, and then conducted threshold segmentation edge effects come to a better place for the segmentation results.)

VC++处理图像模糊或柔化的效果示例源程序，类似Photoshop中的图像高斯模糊效果，每点击一次鼠标右键，程序为模糊一次，直到你满意为止。注意程序仅实现了模糊的功能，其它功能比如程序优化和效率方面的考虑还比较少，因此程序运行速度方面可能会欠缺一些(VC++ deal with image blur or soften the effect of sample source code, similar to Photoshop' s image Gaussian Blur effect, each click the right mouse button, the procedures for fuzzy time until you are satisfied. Note that the program is only vague functions, other functions such as process optimization and efficiency considerations is still relatively small, the speed of the program is running may lack some of the)

说明：  小波降噪，运用小波工具将图像进行分层处理，可以得到降噪后的图像，含有凯斯西储大学轴承数据(Wavelet de-noising: the image is layered by using wavelet tools, and the denoised image can be obtained)