ica_example

内容包括高斯混合概率假设密度滤波器仿真程序，GM-PHD滤波器提供了一种PHD滤波器工程化的方法，在无数据关联的情况下进行目标跟踪。(This algorithm is extended to accommodate mildly nonlinear target dynamics using approximation strategies from the extended and unscented Kalman filters.)

用粒子滤波器实现的多运动员跟踪，包含了data文件(Using Particle Filters athletes realize the multi-tracking, contains a data file)

图像处理 清晰化 .........................(Clarity of image processing .........................)

基于SIFT的一种新特征点匹配算法,比SIFT还强,(Based on SIFT feature points of a new matching algorithm is also stronger than the SIFT,)

说明：  matlab最小二乘法曲面拟合程序可以得到函数具体解析式(Matlab least squares surface fitting program can get the specific analytic expression of the function)

说明：  进行三维去高后，根据sar参数信息，将斜距转变为地距。(After performing three-dimensional height removal, the slope distance is converted to ground distance according to the sar parameter information.)

说明：  dct-svd域水印算法,实现数字水印图像在载体DCT-SVD域的隐藏(a kind of watermarking algorithm in DCT-SVD domain)

自己编的改进谐波小波包实现程序，实现了信号经过谐波小波包变换后幅值不变，适合于机械故障特征提取(The own series improved harmonic wavelet packet achieve signal amplitude unchanged after harmonic wavelet packet transform, for mechanical fault feature extraction)

基于高斯混合模型markov树算法的图像分割。。。(Gaussian mixture model based markov tree algorithm for image segmentation. . .)

稀疏矩阵的实例，讲述如何用稀疏矩阵进行实际的编程。本例是以超市购物为实际背景。(Examples of sparse matrix)