IntDCT

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)?).

基于论文<<Blur Detection for digital images using wavelet transform>>,使用Haar小波变换实现图像模糊度检测。(Based on paper <Blur Detection for digital images using wavelet transform> > , Using the Haar wavelet transform of image blur detection.)

its c-wavelet transform

基于DB4 的小波变换，实现了小波的分解与重构，可根据实际需要修改阈值的选取方法(Based DB4 wavelet transform to achieve wavelet decomposition and reconstruction, according to the actual need to modify the threshold selection method)

说明：  小波变化代码，是比较好的测试向量，包括位平面编码和MQ(Wavelet transform code, is a better test vectors, including the bit plane coding and MQ)

去除在心电信号采集过程中混入的肌电干扰、工频干扰、基线漂移等噪声信信号，避免噪声对心电信号特征点的识别与提取造成误判漏判 已通过测试。 (Remove the EMG interference mixed with the signal acquisition process in mind, frequency interference, baseline drift and noise channel signal to avoid noise caused by the misjudgment of the Missing has been tested on ECG feature identification and extraction.)

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.)

任意阶daubechies小波函数的构造，包括尺度函数和小波函数以及滤波器系数的构造(arbitrary order PG wavelet function, including scaling function and wavelet function and the structure of the filter coefficients)

小波分解在图像处理和电能质量分析中应用广泛，本文应用小波分解的方法将高频的暂态信号提取出来，获取暂态信号的特征量，就能够有效地计算暂态信号。(Wavelet decomposition in image processing and power quality analysis is widely used, this application of wavelet decomposition method to extract high-frequency transient signals, to obtain the amount of transient characteristics of the signal, we can effectively calculate the transient signal.)

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)?).