基于多尺度主成分分析的地震数据局部斜率的计算

发布时间：2018-02-21 08:07

本文关键词： 数据斜率主成分分析多尺度分解　出处：《哈尔滨工业大学》2015年硕士论文　论文类型：学位论文

【摘要】：地震勘探是地球物理研究的重要途径之一。传统的地震数据分析主要是在预处理数据的基础上,建立合适的速度模型,通过精确的速度信息进行数据处理。然而,精确的速度求取,不仅需要花费大量的时间,而且涉及的参数不稳定,使得数据处理结果存在一定偏差。而数据的斜率属性几乎包含了数据的全部信息,并且容易求取,因此,基于局部斜率的数据处理方法在地震勘探中发挥了越来越重要的作用。本文提出了一种基于多尺度分解和主成分分析(PCA)原理的地震反射波同向轴斜率计算的方法。该方法通过图像的高斯金字塔分解,将图像分解到不同分辨率上,然后,进行各尺度图像边缘提取和图像梯度的计算,并对各个尺度的图像梯度进行相同模式和数量的分块。由于主成分分析方法和奇异值分解(SVD)的等价性,我们采用SVD方法简便得到数据块的局部方向的极大似然估计,并通过不同尺度间信息的递归,将低分辨率斜率信息传递到高分辨率图像层。在整个的计算中,通过不同尺度信息的传递减少了噪声对图像数据斜率计算精度的影响,得到比较精确地斜率信息。此外,本文将分别实现不含噪声、含有噪声的合成地震数据及真实数据进行斜率计算的数值模拟,以证明该方法的可行性、对噪声的鲁棒性。
[Abstract]:Seismic exploration is one of the important ways in geophysical research. Traditional seismic data analysis is mainly based on preprocessing data, establishing appropriate velocity model and processing data by accurate velocity information. It not only takes a lot of time to calculate the accurate speed, but also the parameter involved is unstable, which makes the data processing result have certain deviation. The slope attribute of the data contains almost all the information of the data, and it is easy to get, therefore, the slope attribute of the data contains almost all the information of the data, so, The data processing method based on local slope plays a more and more important role in seismic exploration. In this paper, a method for calculating coaxial slope of seismic reflection wave based on the principle of multi-scale decomposition and principal component analysis (PCA) is proposed. The method is decomposed by Gao Si pyramid of the image, The image is decomposed into different resolutions, and then the image edge extraction and image gradient are calculated at various scales. Because of the equivalence of principal component analysis (PCA) and singular value decomposition (SVD), we use the SVD method to obtain the maximum likelihood estimation of the local direction of the data block. The low resolution slope information is transferred to the high resolution image layer through the recursion of the information between different scales. In the whole calculation, the influence of noise on the accuracy of slope calculation of image data is reduced by the transmission of different scale information. In addition, the slope of synthetic seismic data without noise and real data are simulated respectively to prove the feasibility of the method and its robustness to noise.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.4;TP391.41

【参考文献】