数字图像在产生过程中经常会受到噪音污染。因此，对于数字图像应该对其进行去噪音处理。在传统的基于傅里叶变换的信号去噪音方法中，可以使得信号和噪音的频带重叠部分尽可能的小，这样就可以在频域通过时，不需改变滤波方法而将信号和噪音分隔开来。但是，如果信号和噪音的频域重合时，用该方法进行噪音处理的效果就比较差。-digital image in the selection process will be subject to regular noise pollution. Therefore, the digital image should be carried out to deal with the noise. Based on the traditional Fourier Transform method of signal to noise, can result in a signal and noise bands overlap small as possible, so it can be adopted in the frequency domain, the need to change methods of filtering the signal and noise to separate. However, if the signal and noise in the frequency domain coincidence, the method for dealing with the effects of noise on the relatively poor.