wavelet_lab2

实现了对图像最基本的分解重构融合的较为简单的程序 (Realize the decomposition of the image reconstruction of the most basic integration of relatively simple procedures)

关于小波变换及其应用的一本经典教材，值得学习(A classic textbook on wavelet transform and its application is worth learning)

利用小波变换模极大值方法进行信号检测的实验,很容易的检测出突变点(Wavelet transform modulus maxima method of signal detection experiments, it is easy to detect mutations)

运行emd,得到输入信号X的固有模态函数（imf），每列代表着信号X的一阶分量，X就是我们采集的信号。 hhtdn是hht去噪的程序，可以不看，仅看emd 和 findpeaks wvtdn是wavelet去噪程序，如果只是分解信号，仅用下面的语句即可 [C,L] = wavedec(x_noise,dl,wn); x_noise就是我们采集的信号，dl是分解的阶数，wn是分解的母小波名称(% Empiricial Mode Decomposition (Hilbert-Huang Transform) % imf = emd(x) % Func : findpeaks)

一种新的瞬时频率估计方法，用到了同步压缩短时傅里叶变换的高阶形式(This MATLAB package is a collection of scripts allowing to generate figures in the paper [1]. This paper puts forward a generalization of the short-time Fourier-based Synchrosqueezing Transform using a new local estimate of instantaneous frequency. Such a technique enables not only to achieve a highly concentrated time-frequency representation for a wide variety of AM-FM multicomponent signals but also to reconstruct their modes with a high accuracy.)

基于小波变换的特征提取，并且给出了利用小波变换的纹理特征提取和颜色特征提取。(Feature extraction based on wavelet transform and wavelet transform are given texture and color feature extraction feature extraction.)

说明：  用MATLAB编程实现实现滤波，与数字信号处理相关(Implementing filtering)

说明：  基本的S变换matlab实现 用于时频分析 为初学者提供基本代码(S transformation for time-frequency analysis to provide basic code for beginners)

说明：  提出了一种结合SVD的小波变换方法，对其在外弹道测量数据中的野值剔除进行了研究。对观测数据进行小波分解，将小波分解后的近似分量和细节分量组合实现相空间重构，作为SVD方法的输入观测矩阵，根据奇异 熵增量准则，对奇异值进行筛选，根据SVD逆变换重构原信号。这一方法克服了Hankel矩阵相空间构建方法数据 端点失真问题。以小波分解后分量重构的相空间可以满足正交性，进一步提高了SVD进行数据降噪和野值检测的精度。仿真数据和试验数据处理结果证明了这一方法的有效性。(Proposed a method of combining the SVD wavelet transform its outer ballistic measurement data in excluding outliers were studied. Observational data on wavelet decomposition, wavelet decomposition of the approximate combined component and detail component to achieve the phase space reconstruction, as the input observation matrix SVD method, based on singular entropy increment standards, singular value filter, according to the inverse transform heavy SVD structure of the original signal. This approach overcomes the Hankel matrix phase space construction method data endpoint distortion. Wavelet decomposition component of reconstruction phase space satisfy orthogonality, to further improve the SVD for data noise reduction and outlier detection accuracy. Simulation data and experimental data processing results prove the effectiveness of this approach.)

说明：  用于计算一个实数或复数序列的离散傅里叶变换和一个复数序列的逆离散傅里叶变换(fast fourier transform subroutine for real valued series)