image

这个matlab程序实现了Edge Linking and Line Segment Fitting，解压后直接运行demo.m就可以了 ( edgelink.m edge linking function that forms lists of connected edge points from a binary edge image. (Needs findendsjunctions and cleanedgelist below). drawedgelist.m plots out a set of edge lists generated by edgelink or lineseg. edgelist2image.m transfers edgelist data back into a 2D image array. lineseg.m forms straight line segments fitted with a specified tolerance to the lists of connected edge points. maxlinedev.m is also used by lineseg.m to calculate deviations of the edge lists from the fitted segments. findendsjunctions.m finds line junctions and endings in a line/edge image. cleanedgelist.m cleans up a set of edge lists generated by edgelink or lineseg so that isolated edges and spurs that are shorter than a minimum length are removed.)

用于盲信号分离的LMS算法，可以从混合的观测向量中分离出需要的信号，判读的标准是计算串音误差(For the LMS algorithm for blind signal separation can be observed from the hybrid vector needs to separate out the signals, the standard interpretation is to calculate the crosstalk error)

Bp神经网络实现语音信号的分类 效果十分好 值得参考(Language signal classification Bp neural network)

dmc的matlab实现方法，运行实现过程正常，中间有一部分注释(matlab code of dmc which have been tested for many times)

SAR雷达成像仿真 基于MATLAB的三种不同的算法程序 分析频谱与分辨率(SAR radar imaging simulation based on MATLAB for three different algorithms of spectrum and resolution procedures)

to get the solution of Orr-Sommerfeld equation

KSW最大熵阈值分割,根据像素分布概率来进行阈值选取(KSW maximum entropy thresholding)

说明：  用MATLAB编写的svm源程序，可以实现支持向量机，用于特征分类或提取(SVM prepared using MATLAB source code, you can achieve the support vector machine for feature classification or extraction)

matlab中如果需要更改对话窗口的图标，则需要自己在程序中加入代码，文件包括3个窗口的图标更改程序。(matlab dialog window, if you need to change the icon, you will need to add your own in the program code, files including three windows icons change the program.)

Bayesian mixture of Gaussians. This set of files contains functions for performing inference and learning on a Bayesian Gaussian mixture model. Learning is carried out via the variational expectation maximization algorithm.