subtraction-speech-denoising

小波分析，及小波能量谱分析，信号分析与处理技术小波分析，及小波能量谱分析，信号分析与处理技术(wavelet analysis, and wavelet energy spectrum analysis, signal analysis and processing technology wavelet analysis, and wavelet energy spectrum analysis, Signal Analysis and Processing Technology)

雷达系统设计部分信号处理程序，包括部分目标和干扰模型、匹配滤波、波形设计、多普勒处理、门限检测及恒虚警率等方面信号处理的程序(Signal processing program for radar system design)

实现了高斯混合尺度模型（GSM）算法，可被嵌入到各种去噪框架中（如各种小波变换等）使用。(Gaussian mixture scale model (GSM) algorithm, can be embedded into a variety of denoising framework (such as a variety of wavelet transform, etc.) use.)

小波阈值去噪的经典程序，基于小波变换阈值去噪的MATLAB源代码，效果很好。 (Classic wavelet threshold denoising procedure threshold denoising based on wavelet transform of MATLAB source code, works well.)

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.)

小波信号分解与重构的Matlab程序，程序里边涉及卷积的直接调用的Matlab的conv()函数。(Wavelet signal decomposition and reconstruction in Matlab,Program inside invoked directly related to the convolution of Matlab conv () function.)

基于小波变换的图像压缩编码，实现了多种小波变换，并且实现了12:1的高压缩率。(Based on Wavelet Transform Image Coding, to achieve a variety of wavelet transform, and achieved a high 12:1 compression ratio.)

its c-wavelet transform

matlab程序源代码,包括小波分析理论与MATLAB R2007实现所有代码。(matlab program source code, including all code wavelet analysis theory and MATLAB R2007.)

移动窗口二维相关光谱方法matlab的实现，完全是自己参照公式编的哦。。移动窗口二维相关光谱是一种新的二维相关分析方法,它将移动窗口的概念和二维相关分析方法有效地结合在了一起,利用移动窗口将庞大的光谱数据按矩阵分割成若干个便于操作的子矩阵,用二维相关光谱分别处理,将得到的结果综合分析,用以光谱变量和扰动变量为坐标的等高图表现出来,从中可以非常直观地观察出光谱强度在扰动变量方向上的变化,进而找出引起光谱强度突变的特征扰动点。(Moving window two-dimensional correlation spectroscopy method matlab realization, is entirely their own formulas reference the Oh. . Moving window two-dimensional correlation spectroscopy is a new two-dimensional correlation analysis method, it will move the window concept and two-dimensional correlation analysis method is effective to combine together, using the moving window at the huge spectral data matrix is ​ ​ divided into a number of easy to operate a sub-matrix, respectively, using two-dimensional correlation spectroscopy process, comprehensive analysis of the results obtained for the spectral variables, and disturbance variables manifested contour map coordinates, which can visually observe the spectral intensity of the direction of the disturbance variable The change, then find mutations characteristic spectral intensity caused disturbance point.)