论文部分内容阅读
针对信号滤波中噪声方差的估计、阈值的确定及细节信息的保留三个问题,利用添加高斯白噪声的模拟信号进行实验分析,结果表明:噪声主要集中在低尺度,因此可以在低尺度空间估计噪声方差;噪声的能量随尺度的增大而逐渐减弱,阈值采用自适应算法确定较为合适;小波自适应滤波残留较多的小噪声,因此提出了尺度相关与小波自适应相结合的滤波算法.该算法的滤波结果表明,信号的信噪比较之传统小波算法可以提高10%.并且由目视效果可以看出,通过该算法滤波后的信号较之传统小波算法更加平滑.
According to the three problems of noise variance estimation, threshold determination and detail information preservation in signal filtering, the experiment is carried out by using analog signals with Gaussian white noise. The results show that the noise is mainly concentrated in low-scale and therefore can be used in low-scale spatial estimation Noise variance. The energy of noise is gradually weakened with the increase of scale, and the threshold is more suitable to be determined by adaptive algorithm. The wavelet adaptive filtering has more residual noise, so the filter algorithm combining scale correlation with wavelet adaptive is proposed. The filtering results of this algorithm show that the signal-to-noise ratio of the signal can be increased by 10% compared with the traditional wavelet algorithm, and it can be seen from the visual effect that the signal filtered by the algorithm is smoother than the traditional wavelet algorithm.