rician

说明：  按FPE定阶的 源程序：fpe.cpp M序列：M序列.txt 白噪声：Gauss.txt 程序中先用依模型阶次递推算法估计模型的参数，再用fpe方法判断模型的阶次。 程序运行结果如下： n: 1 判断阶次FPE的值: 0.0096406 -0.481665 1.07868 n: 2 判断阶次FPE的值: 0.00875755 -0.446739 0.00498181 1.07791 0.0527289 n: 3 判断阶次FPE的值: 0.0087098 -0.459433 0.120972 -0.0569228 1.07814 0.0390757 0.116982 n: 4 判断阶次FPE的值: 0.000396884 -0.509677 0.4501 -0.200906 0.0656188 1.07991 -0.0156362 0.442989 0.0497236 n: 5 判断阶次FPE的值: 3.2095e-007 -1.18415 0.813123 -0.517862 0.34881 -0.116864 1.07999 -0.744141 0.474462 -0.253112 0.122771 n: 6 判断阶次FPE的值: 3.23349e-007 -1.14659 0.76933 -0.487651 0.329676 -0.10377 -0.00440907 1.07999 -0.703574 0.447253 -0.235282 0.113587 0.00479688 从以上结果可以看出，当n=5时，fpe值最小，所以这时的模型阶次和参数估计值为最优结果： 3.2095e-007 -1.18415 0.813123 -0.517862 0.34881 -0.116864 1.07999 -0.744141 0.474462 -0.253112 0.122771(err)

说明：  用函数graycoprops去检测图像灰度共生矩阵，该例子主要是应用于图像纹理检测(Use function graycoprops to check picture ,it is very good )

Gaussian process latent variable model code for dynamic texture synthesis

Multiple instance metric learning automatically labeled bags of faces

说明：  finds polynomials using spline algorithm

是支持向量机工具箱（matlab），可用于模式分类，模式识别等。(support vector machines Toolbox (Matlab), can be used for pattern classification, pattern recognition.)

可以把输出图片设置为任意格式 MATLAB程序(Exports figures nicely to a number of vector & bitmap formats)

fcm算法的matlab实现，6.5以上版本可用(FCM algorithm matlab realized, more than 6.5 version is available)

说明：  实现MATLAB中解决基线漂移的算法，使图线更便于观察(MATLAB implementation of the algorithm to solve the baseline drift, so that more easy to observe Chart)

绝对的lowe老人家提出的sift算法源码，配套有文档说明，对于初学者极度有用，还有我添加的注释我移植的仿射变换函数，绝对值得下的资源~！(The absolute lowe elderly sift algorithm proposed source code, supporting have documented for beginners extremely useful, and I add comments I transplanted affine transformation function, absolute worthy resource ~!)