包括一次行、列变换，一层，二层，三层小波变换，低通，高通滤波，图像复原，增强，融合...

阐述访生小波变换的理论及在心电信号去噪中的应用-Visit the Health and wavelet transform theory and the heart signal denoising

Contourlet变换的实现源码。相关理论可参考:Do M N, Vetterli M. "The contourlet transform: an efficient directional multiresolution image representation" [J]. Image Processing, IEEE Transactions on, 2005, 14 (12): 2091-2106.-Transform realize Contourlet source. Relevant theory may refer to: Do MN, Vetterli M. The contourlet transform: an efficient directional multiresolution image representation [J]. Image Processing, IEEE Transactions on, 2005, 14 (12): 2091-2106.

小波神经网络程序基于反向传播训练-wavelet-based neural network program back propagation training

基于小波变换的彩色图像压缩.-based on wavelet transform color image compression.

快速傅立叶变换代码 快速傅立叶变换代码-Fast Fourier Transform Fast Fourier Transform source code

详细介绍了小波变换的相关知识-Detailed knowledge of the wavelet transform

一个时频分析的工具箱,主要进行小波变换的信号处理函数-A time-frequency analysis of the toolbox, the main wavelet transform signal processing function

小波变换具有良好的时―频局部性，是分析奇异信号的重要方法定点DSP在工程中的厘用士分背通，具实现小波变换可以满足工程是实时性的要求文中简要介绍了小波变换理论及算法，并结合TI公司的16位定点说明算法的实现。 -Wavelet transform has a good time- frequency localization, is to analyze the singular signal an important means of fixed-point DSP in engineering per cent pass with disabilities at the back with the realization of wavelet transform to meet the project is a real-time requirements of the article a brief introduction of wavelet transform theory and algorithm, combined with TI

关于二维小波变换的程序 [精华] 说明：此算法重在概念，速度并不是很快。因为FOR循环的缘故。此程序从循环矩阵的观点出发，把圆周卷积和快速幅里叶变换建立了联系。实现了分解和无失真重构。它只做了一层分解，即将256x256图形分解成为64x64的四个图形，避免了使用WKEEP()的困惑。主要思想为用小波滤波器族构造正交阵W,变换写为B=W*A*W ,反变换为：A=W *A*W,这与所有正交变换无异。W为循环正交矩阵，因此可用FFT实现快速运算，难点就在重构矩阵上。若用矩阵概念明确，一个共扼转置可以搞顶。但FFT的使用必须找到与分解序列的关系。-on Wavelet Transform procedures [best] Note : This algorithm important concept is not speed quickly. Because FOR cycle nowadays. This program cycle from the point of view of Matrix, circular convolution and fast Fourier transform is established links. To achieve the decomposition and reconstruction without distortion. It only had one decomposition, 256x256 graphics will be 64x64 decomposition of the four graphics, avoiding the use of WKEEP () confusion. The main idea of using wavelet filter generator orthogonal array W, write to transform B = W* A* W, anti-Transform : A = W* A* W, which is orthogonal transformation all the same. W cycle orthogonal matrix, can be used to achieve rapid FFT computation, on the dif

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