摘要
传统地震数据重建算法大多采用整体重建模式。受数字图像重建思路启发,提出了一种基于高阶扩展快速行进法的缺失地震数据重建算法。该算法采用局部重建模式,首先将缺失地震数据映射为地震图像,并定量分析映射导致的量化误差;再采用二抽取小波变换对地震图像进行分解,分解后的低频分量采用高阶扩展快速行进法做局部逐点重建,高频分量通过已重建低频部分的水平、垂直和对角预测滤波做重建;然后采用小波逆变换得到重建后的地震图像;最后将地震图像映射回地震数据。叠前和叠后实际地震数据重建实例验证了算法的可行性;与基于形态分量分析、基于K-奇异值分解(SVD)字典学习等地震数据重建算法的对比结果表明,本文算法具有更快的重建速度和更高的重建精度。
Most conventional seismic data reconstruction algorithms use the overall reconstruction mode.Inspired by digital image reconstruction,we propose in this paper a novel seismic data reconstruction algorithm based on the high-order expansion fast marching method.The algorithm uses local reconstruction mode.Firstly,the missing seismic data is mapped into seismic image and the quantization error in the mapping is analyzed.Then the seismic image is subsequently decomposed by wavelet transform with two down-sampling.Decomposed low frequency part is reconstructed point by point with the high-order expansion fast marching method.High frequency part is reconstructed based on the horizontal,vertical and diagonal prediction filtering of the low frequency part.After that the reconstructed seismic image is obtained by the inverse wavelet transform.Finally,the seismic image is mapped into the seismic data.Prestack and poststack seismic data reconstruction experiments verify the feasibility of the proposed algorithm.In comparison with morphological component analysis(MCA)and the K-SVD dictionary learning,the proposed algorithm is faster and more accurate in data reconstruction.
出处
《石油地球物理勘探》
EI
CSCD
北大核心
2015年第5期873-880,803,共8页
Oil Geophysical Prospecting
基金
国家自然科学基金(60972106
51475136)
河北省自然科学基金(F2013202254)
中国博士后科学基金(2014M561053)
天津自然科学基金(12JCYBJC12400)联合资助
关键词
地震数据重建
快速行进法
小波变换
数字图像重建
seismic data reconstruction,fast marc-hing method,wavelet transform,digital image re-construction
