基于小波变换的低场脉冲核磁共振信号滤波

Noise Reduction Based on Wavelet Transform for Low-field Pulsed NMR Signal

低场脉冲核磁共振信号在去除噪声的同时不能损失峰值与边沿有用信息。本文首先介绍了基于平稳小波变换的空域阈值滤波法,然后提出对小波系数先平移以保证边沿对齐再做相邻尺度乘积的改进,并给出新的空域相关阈值计算公式,试验表明该方法在有效保护核磁共振信号峰值与边沿的同时很好地抑制了噪声。

It is important to reduce the noise from the low-field pulsed NMR signal while keeping the peaks and edges well. The spatially selective noise filter based on stationary wavelet transform that uses spatial correlation thresholding is firstly introduced. A new method is presented by translating and lining up the edges before multiplying the adjacent wavelet scales. We also propose a new spatial correlation threshold in our scheme. The result of experiment indicates that our method can keep the peaks and edges of the NMR signal well while getting rid of noise greatly.