xiazai2

随机场的估计和应用，是非常流行的图像分割的算法。(markov random field is very popular)

包括IHS遥感图像融合算法，比较全。基于MATLAB。(IHS image merging)

基于高斯混合模型(GMM)的em算法，给出了其matlab实现(em algorithm for GMM)

说明：  通用的空间矢量脉宽调制 SPACE VECTOR PULSEWIDTH MODULATION (SVPWM)源程序(SVPWM based on C language)

Matlab_jiaoan：这是一个matlab教案(Matlab_jiaoan : This is a Matlab templates)

emd算法 主要用来提取信号的本征模态，将信号的原始信号从混合噪声中提取出来(emd algorithm Mainly used to extract the signal intrinsic mode, the original signal from the mixed noise extraction)

一个有趣的桌面撞球模拟,可录制avi,输出坐标参数及图象(An interesting desktop billiards simulation, can record avi, output parameters and image coordinates)

说明：  可以进行傅立叶变换的matlab程序 保证可以使用 (Fourier Transform can guarantee that you can use matlab program)

MUSIC算法是一种基于矩阵特征空间分解的方法。从几何角度讲，信号处理的观测空间可以分解为信号子空间和噪声子空间，显然这两个空间是正交的。信号子空间由阵列接收到的数据协方差矩阵中与信号对应的特征向量组成，噪声子空间则由协方差矩阵中所有最小特征值（噪声方差）对应的特征向量组成。MUSIC算法就是利用这两个互补空间之间的正交特性来估计空间信号的方位。噪声子空间的所有向量被用来构造谱，所有空间方位谱中的峰值位置对应信号的来波方位。MUSIC算法大大提高了测向分辨率，同时适应于任意形状的天线阵列，但是原型MUSIC算法要求来波信号是不相干的。(MUSIC algorithm is a matrix decomposition method based on feature space. From the geometric point of view, the observed spatial signal processing can be decomposed into signal subspace and noise subspace, it is clear that the two spaces are orthogonal. The signal received by the array to the subspace of the covariance matrix of the signal component corresponding eigenvectors, the noise subspace of the covariance matrix of all by the smallest eigenvalue (noise variance) eigenvectors corresponding to the composition. MUSIC algorithm is the use of orthogonal properties between these two complementary space to estimate the spatial orientation of the signal. Noise subspace of all vectors are used to construct the spectrum, all the spatial orientation of the spectrum corresponding to the peak position of the signal wave direction. MUSIC algorithm greatly improves the resolution measurements, while the antenna array adapted to any shape, but the prototype MUSIC algorithm requires to wave sign)

he basic idea of the method of bisection is to start with an initial interval, [a0,b0], that is chosen so that f(a0)f(b0) < 0. (This guarantees that there is at least one root of the function f(x) within the initial interval.) We then iteratively bisect the interval, generating a sequence of intervals [ak,bk] that is guaranteed to converge to a solution to f(x) = 0.