提高地震剖面图像信噪比的二维沿层滤波方法
|
摘要
为了抑制噪声,通过地震数据特征的理论分析和实验研究,在数字图像处理常用的均值滤波方法中引入了相干技术,导出一种提高地震剖面图像信嗓比(SNR)的新方法———二维沿层滤波,并且建立了其数学模型.实验结果分析比较表明,该方法去噪效果明显优于均值滤波,在信号能量保持大于80%的条件下,可使信噪比增加3～8dB,并且可以保持原来的信号能量关系,使地震剖面图像的质量明显改善,有利于进一步处理和地震解释.
Through theory analysis of the seismic data characters and the experimental research in suppressing noise of the seismic image, the method of mean value filtering which is often applied in digital image processing is modified by importing the correlation technique. The novel method of increasing signal to noise ratio (SNR) of seismic section images named 2-D tracing horizon filtering is derived, and the mathematic models of 2-D tracing horizon filtering are established. The analyses and comparisons of experimental results indicate that the effect of this novel method exceeds evidently the one of mean value filtering. It can make signal to noise ratio increase 3～8 dB, while not changing original signal energy relation under the condition of keeping signal energy more than 80%, so the quality of seismic section images can be improved evidently. Therefore, 2-D tracing horizon filtering is available to next processing step and seismic interpretation.
Through theory analysis of the seismic data characters and the experimental research in suppressing noise of the seismic image, the method of mean value filtering which is often applied in digital image processing is modified by importing the correlation technique. The novel method of increasing signal to noise ratio (SNR) of seismic section images named 2-D tracing horizon filtering is derived, and the mathematic models of 2-D tracing horizon filtering are established. The analyses and comparisons of experimental results indicate that the effect of this novel method exceeds evidently the one of mean value filtering. It can make signal to noise ratio increase 3～8 dB, while not changing original signal energy relation under the condition of keeping signal energy more than 80%, so the quality of seismic section images can be improved evidently. Therefore, 2-D tracing horizon filtering is available to next processing step and seismic interpretation.
引文
[1]PETERSONJB,DEGROATRD.NEARLYOPTIMAL,SVD-BasedStackingofSeiusmicData[M].MAPLE PRESS,1988.858-862.
[2]刘振东,何和英.外推算子预测去噪方法[J].石油物探,2001,40(1):26-32.
[3]BAISHALIR,MEREURF.Signalenhancementusing patternrecognitiontechniqueswithapplicationtonear verticalcrustalseismicreflectionexperiments[J].Geo physcalResearcgKetters,1996,23(14):1849-1852.
[4]BONDARI.Seismichorizondetectionusingimagepro cessingalgorithms[J].GeophysicalProspecting,1992,40:785-800.
[5]付雷,田晓冬.相干体技术在三维地震资料中的应用[J].世界地质,2000,19(3):279-281.
[6]程荃,黄先律.提高迭后地震记录分辨率的频率振幅补偿方法[J].物探化探计算技术,1996,18(3):238-241
[2]刘振东,何和英.外推算子预测去噪方法[J].石油物探,2001,40(1):26-32.
[3]BAISHALIR,MEREURF.Signalenhancementusing patternrecognitiontechniqueswithapplicationtonear verticalcrustalseismicreflectionexperiments[J].Geo physcalResearcgKetters,1996,23(14):1849-1852.
[4]BONDARI.Seismichorizondetectionusingimagepro cessingalgorithms[J].GeophysicalProspecting,1992,40:785-800.
[5]付雷,田晓冬.相干体技术在三维地震资料中的应用[J].世界地质,2000,19(3):279-281.
[6]程荃,黄先律.提高迭后地震记录分辨率的频率振幅补偿方法[J].物探化探计算技术,1996,18(3):238-241
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心