摘要
传统广义希尔伯特变换在处理弱振幅分布较广或者有地质褶皱带的复杂地震数据时,强振幅周围的边缘检测效果还相对比较清晰,但弱振幅周围的边缘检测就十分模糊,同时细节信息显示不够全面。因此,该算法在Luo等人提出的广义希尔伯特变换算法的基础上进行优化,提升弱振幅周围和地质褶皱带的边缘检测效果。首先,将地震信号在时间域上加高斯窗并做傅里叶变换;然后取傅里叶变换后信号的虚部,取绝对值,沿频率求和后计算1/n次幂;最后,沿纵轴和横轴分别进行边缘检测,将两者的结果求取平均作为检测结果。该算法应用于实际地震数据时,弱振幅和地质褶皱带周围的图像边缘被清晰地检测出来且计算效率明显快于频域加窗的广义希尔伯特变换,验证了文章提出的算法的有效性。
When the traditional generalized Hilbert transform processes complex seismic data with a wide range of weak amplitudes or geological folds, the edge detection effect around strong amplitude is still relatively clear, but the edge detection around the weak amplitude is very vague, and details Information display is not comprehensive enough.Therefore, the algorithm is optimized based on the generalized Hilbert transform algorithm proposed by Luo et al. to improve the detection effect of the edges around the weak amplitude and the geological fold belt.First,the seismic signal is added to the Gaussian window in the time domain and the Fourier transform is performed.Then take the imaginary part of the signal after Fourier transform, take the absolute value, calculate the 1/n power after summing the frequencies;Finally, edge detection is performed along the vertical axis and the horizontal axis respectively, and the results of both are averaged as the detection result.When the algorithm is applied to actual seismic data, the edges of the images around weak-amplitude and geological folds are clearly detected and the computational efficiency is significantly faster than the frequency-domain windowed generalized Hilbert transform. The validity of the algorithm proposed in this article has been verified.
作者
常强
ChangQiang(College of communication engineering,Chengdu University of Information Technology,Cheng du,China,610225)
出处
《信息通信》
2018年第9期10-12,共3页
Information & Communications
