TAnalysis

对于PCA ，LDA，BIC，EBGM都有很详细的代码。但是需要在LINUX下配置运行。不过代码较有参考价值。(For PCA, LDA, BIC, EBGM have a very detailed code. But the need to configure to run under LINUX. However, a more useful code.)

小波变换模极大值重建信号的matlab源程序，来源于胡广书的《现代信号处理教程》附属光盘(Wavelet transform modulus maxima reconstruction signal matlab source code, derived from HU-book " modern signal processing tutorial subsidiary Disc)

在理解了离散小波变换的基本原理和算法的基础上，通过设计VC程序对简单的一维信 号在加上了高斯白噪声之后进行Daubechies小波、Morlet小波和Haar小波变换，从而得到小波分解系数；再通过改变分解得到的各层高频系数进行信号的小波重构达到消噪的目的。在这一程序实现的过程中能直观地理解信号小波分解重构的过程和在信号消噪中的重要作用，以及在对各层高频系数进行权重处理时系数的选取对信号消噪效果的影响。(In understanding the basis of discrete wavelet transform basic principles and algorithms, through the design VC program Daubechies wavelet, Morlet wavelet and Haar wavelet transform after a simple one-dimensional signal plus a Gaussian white noise, resulting wavelet coefficients wavelet reconstruction signal again by changing the layers to achieve high-frequency coefficients decomposed noise cancellation purposes. Can intuitively understand wavelet decomposition and reconstruction process and an important role in signal de-noising in the process of implementation of this program, and in the high-frequency coefficients when the weight of the layers selected for signal processing coefficients denoising affected.)

基于小波变换的图片重建 有完整代码 工程 以及说明 涉及到图片的切割(image reconstruction based on wavelet translate )

实现离散余弦(TCD)变换(achieve discrete cosine (TCD) Transform)

说明：  GHM多小波的变换与逆变换，包括预处理和后处理(GHM multi-wavelet transform and inverse transform)

具有提升分解结构的整数离散变换，可用于基于提升格式的小波变换设计。(Enhance the breakdown structure of the integers with the discrete transform can be used for wavelet transform based on lifting scheme design.)

此程序实现图象傅立叶变换、离散傅立叶变换和小波变换、可以自己设定小波变换的参数(this procedure image Fourier transform, discrete Fourier transform and wavelet transform, can set their own parameters of wavelet transform)

Wavelet compression algorithm. Start by computing a complete wavelet-transform using the cohen-daubechies-fevereau (2,2) biorthogonal wavelet (which incidentally seems to be the same as used for lossless JPEG2000, but they call it a (5,3)?).

Wavelet compression algorithm. Start by computing a complete wavelet-transform using the cohen-daubechies-fevereau (2,2) biorthogonal wavelet (which incidentally seems to be the same as used for lossless JPEG2000, but they call it a (5,3)?).