基于小波变换分割图像

dual tree complex wavelet transform

说明：  曲波变换和去噪，可以进行曲波变换阈值去噪，是很好的去噪入门代码。(Curvelet transform and denoising, can be threshold denoising curvelet transform, is a good denoising entry code.)

针对转子系统非线性动力学图谱分析，绘制转子的轨迹图，相图，频谱图以及庞加莱图。(analysis about nonlinear dynamics of the rotor bearing system)

与labview之间的联系 在小波变换方面(Contact between labview in terms of the wavelet transform)

现代控制理论，卡尔曼滤波器设计，simulink控制仿真(Modern control theory, Calman filter design, Simulink control simulation.)

小波神经网络,实现曲线拟和，数据分类，故障诊断，学习速度快于传统BP。(Wavelet neural network, and realize curve fitting, data classification, fault diagnosis, learning faster than the traditional BP.)

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.)

说明：  很详细的hht编程代码，matlab亲测可用，提供了比较清晰的注释，帮助理解(Very detailed HHT programming code, matlab pro test available, provides a relatively clear note to help understand)

基于小波变换的随机噪声降噪的matlab 主要运用的方法有软阈值去噪法和硬阈值去噪法(Based on wavelet transform matlab random noise and noise reduction methods are mainly used soft thresholding method and hard thresholding method)

旨在用C语言程序和MATLAB程序通过MUSIC算法对“信号提取”进行实现，结果证明程序运行效果良好，分辨率较高(Designed with C language program and MATLAB program through the MUSIC algorithm " signal extraction" for implementation, the results show the program running well, higher resolution)