LMMSE

利用hough变换检测圆，有RHT，随机hough变换，以及一些改进(Using Hough transform to detect the circle)

滤波反投影CT图像重建算法 filter-backproject reconstruction, including feldkamp (FDK) cone beam(Filter back projection algorithm for CT image reconstruction filter-backproject reconstruction, including feldkamp (FDK) cone beam)

两个l1准则下的噪声干扰信号压缩感知重构举例，两个例子的稀疏矩阵均为DCT矩阵，而观测矩阵分别采用单位阵和随机矩阵，有详细的步骤和使用方法，适用于初步的学习压缩感知方法。(This programme supply two examples by Compressed sensing with l1 norm. The sparse matrix of two examples are all DCT matrix and the obsever matrix are unit matrix and random matrix respectively. We give the detailed usage and steps for the programme which is useful to the beginner.)

使用经典边缘检测算子进行边缘检测，加入噪声的图像边缘检测结果和未加噪声的图像检测结果进行对比分析，默认阈值设置的图像检测结果和手动设置阈值的图像检测结果进行对比(Edge detection using classical edge detection operator)

应用于fluent的UDF程序，用于修正液滴表面轮廓(The UDF is applied fluent procedures for correcting the droplet surface contour)

主要思想是用meanshift算法跟踪目标，拖动矩形框选择目标模板，meanshift自动跟踪 本程序的主要缺陷体现在帧差法自动识别目标，这部分功能与上一个功能独立，可用可不用，用于自动识别目标，效果不是很好，主要体现在对图像视频的要求是 背景（目标以外的所有事物）不能动(The main idea is to use meanshift algorithm to track the target and drag a rectangle to select the target template, meanshift automatic tracking The main drawback of the program is reflected in the frame difference method automatically identifies the target, which is part of the function and one function independently, may or may not, be used to automatically identify the target, the effect is not very good, mainly reflected in the requirements of the video image is the background (target All things outside) can not move)

雷达信号处理一些相关操作，计算其距离向，方位向以及多普勒参数等。(radar signal processing)

SRCNN Superresolution imteplated by tensorflow SRCNN tensorflow 实现(SRCNN Superresolution imteplated by tensorflow)

利用剪切波变换的方向性和多尺度特性，在剪切波变换域中进行阈值降噪处理(Using the directional and multi-scale characteristics of shearlet, the threshold denoising is performed well in the shearlet transform domain)

image decomposition and reconstruction