Elmane_modeling

这个是用Matlab写成的说话人识别演示程序。(Currently, there are more than 20 classes in this library, including commonly used feature extraction algorithms and modeling techniques for speech recognition and speaker verification.-Currently, there are more than 20 classes in this library. including commonly used feature extraction al gorithms and modeling techniques for re speech cognition and speaker verification.)

clear all I=imread( lena.bmp ) figure imshow(I) I2=imrotate(I,-4, bilinear ) 逆时针旋转4度 figure imshow(I2) I3=fliplr(I) 垂直镜像 figure imshow(I3) I4=imresize(I,0.5, bilinear ) 缩小为原图的1/2 figure imshow(I4) A=double(I) 转换为double类型 计算7个不变矩 [nc,nr]=size(A) [x,y]=meshgrid(1:nr,1:nc) 得到网格 x=x(:) y=y(:) A=A(:) m.m00=sum(A) if m.m00==0 m.m00=eps end m.m10=sum(x.*A) m.m01=sum(y.*A) 计算均值 xmean=m.m10/m.m00 ymean=m.m01/m.m00 计算中心矩(li9_23.rar cm.cm00=m.m00 cm.cm02=(sum((y-ymean).^2.*A))/(m.m00^2) cm.cm03=(sum((y-ymean).^3.*A))/(m.m00^2.5) cm.cm11=(sum((x-xmean).*(y-ymean).*A))/(m.m00^2) cm.cm12=(sum((x-xmean).*(y-ymean).^2.*A))/(m.m00^2.5) cm.cm20=(sum((x-xmean).^2.*A))/(m.m00^2) cm.cm21=(sum((x-xmean).^2.*(y-ymean).*A))/(m.m00^2.5) cm.cm30=(sum((x-xmean).^3.*A))/(m.m00^2.5) im(1)=cm.cm20+cm.cm02 im(2)=(cm.cm20-cm.cm02)^2+4*cm.cm11^2 im(3)=(cm.cm30-3*cm.cm12)^2+(3*cm.cm21-cm.cm03)^2 im(4)=(cm.cm30+cm.cm12)^2+(cm.cm21+cm.cm03)^2 im(5)=(cm.cm30-3*cm.cm12)*(cm.cm30+cm.cm12)... *((cm.cm30+cm.cm12)^2-3*(cm.cm21+cm.cm03)^2)... +(3*cm.cm21-cm.cm03)*(cm.cm21+cm.cm03)... +(3*(cm.cm30+cm.cm12)^2-(cm.cm21+cm.cm03)^2) im(6)=(cm.cm20-cm.cm02)*((cm.cm30+cm.cm12)^2-(cm.cm21+cm.cm03)^2)... +4*cm.cm11*(cm.cm30+cm.cm12)*(cm.cm21+cm.cm03) im(7)=(3*cm.cm21-cm.cm03)*(cm.cm30+cm.cm12)... *((cm.cm30+cm.cm12)^2-3*(cm.cm21+cm.cm03)^2)... +(3*cm.cm12-cm.cm30)*(cm.cm21+cm.cm03)... *()

说明：  本文介绍了模拟退火算法的原理和求解方法,并将其用于指数曲线的拟合,在Matlab语言环境下实现了该算法,并且与文献中的遗传算法、线性回归相比较。数值模拟结果表明,该算法能更好地实现最优拟合。(This article describes the principle of simulated annealing algorithm and solution methods, and used an exponential curve fitting, the language environment in Matlab implementation of the algorithm and the genetic algorithm with the literature, linear regression comparison. Numerical simulation results show that the algorithm can better achieve the optimal fit.)

MATLAB典型实验和例子，包含程序，供初学者使用，简单易学(MATLAB and examples of typical experiments, including the program for beginners to use, easy to learn)

这是一个基于MATLAB编写的BUTTERWORTH滤波器的程序.其中从s域到z域的变换是自己编写的,代替了MATLAB中的系统BUTTERWORTH的系统函数(对于毕业设计的学生,比较有用!!)(based on MATLAB BUTTERWORTH prepared by the filter process. From which z's domain to the domain transform their own preparation, instead of MATLAB system BUTTERWORTH system function (for graduate design students more useful! !))

利用matlab写的高斯消元法和最速下降法程序(zsxjf base on matlab)

脉搏波的数据以及特征提取，提取到最高点最低点，周期。可以看下。(Pulse wave data, and feature extraction, extract the lowest point to the highest point, cycle. You can have a look.)

LTE咬尾卷积编码，C55上验证过，回环测试成功。(LTE tail conv coder， test passed！)

现代通信系统 一书的全部matlab源码(Modern communication systems all the source code of a book)

一种改进的混沌算法，结合进化算法进行改进,利用logistic映射进行改进(An improved chaotic particle algorithm to improve the diversity of PSO)