超声检测中维纳逆滤波解卷积方法的改进研究

维纳逆滤波是利用解卷积技术提高超声检测纵向分辨率的算法之一。本文分析了其传统算法 对噪声适应性不理想的原因,提出一种改进方法,利用检测系统信噪比先验知识,进行频域解卷积处 理,在此基础上,对检测系统频带内的处理结果建模,插补估计出频带外受噪声干扰影响较大的数 据,然后再变换到时域。计算机模拟及相应实验结果表明,这些改进方法较为理想,具有较强的噪声 适应能力。

A modified Wiener inverse filter for deconvolution in ultrasonic detection

Wiener inverse filter is one of the deconvolution methods to renew echo for improving the vertical resolution. This paper explains why the traditional algorithm is not stable, and then presents a modified algorithm to improve the problem, which is based on using a modified prior SNR and applying the spectrum extrapolation method to the date outside of the frequency band. Computer simulation and experimental observation show that this method of modification is feasible.