利用SVD法重建地震剖面

SEISMIC SECTION RECONSTITUTION BY SVD TECHNIQUE

地质矿产部石油物探研究所

矩阵的奇异值分解法(即SVD法)可用于地震资料处理。地震剖面作为一个二维矩阵,可分解为一系列子特征图象。利用这些子特征图象对原地震剖面作不同阶数的重建,即可获得保留了原地震剖面主要有效特征的新的剖面图象。本文简述了SVD分解的数学原理和特征成像的方法,并分别对理论模型和实际资料进行了试处理。结果表明,这种方法对改善剖面质量和提高信噪比是有效的。

The singular value decomposition technique of the matrix (e. g. ,the SVD technique) might be used for seismic data processing. The seismic section,as a 2-D matrix,could be decomposed into a series of sub-characterized images. By using such images,the original seismic section could be reconstituted with any order,in this way,a new section with the original character will thus be obtained.In this paper,the mathematical principle of the SVD technique and the characteristic imagery method were described in brief, the trial processing was made for the hypothetical model and the actual data. Results show that in improving the quality of the section and the S/N ratio,the technique recommended is indeed very effective and helpful.