摘要
在地震勘探中 ,经过各种处理后最终得到用于解释的地震剖面实际上都可以看作是一些图像。这样 ,就可以从图像处理的观点来看问题 ,图像的特征有边缘和区域 ,那么 ,地震反射同相轴在地震剖面图像上 ,可以看作是一边缘。人工神经网络是近年来发展起来的一种新技术 ,它可以进行二维图像的边缘检测 ,基于这一点 ,把神经网络边缘检测技术应用到地震剖面上来检测地震反射界面 ,压制地震剖面上的随机干扰 ,从而改善地震图像的质量 ,达到提高地震资料信噪比和分辨率的目的。运用该技术对实际剖面进行了处理 ,效果很好 ,证明了方法的可行性和有效性。
In seismic exploration, seismic sections obtained by various processing techniques and used for seismic interpretation are actually regarded as some images. Therefore, we can study the problem from the viewpoint of image processing.An imsge has a region and edges, so a seismic event can be considered as edges on the seismic image. Artificial neural network is a new technique which can make edge detection of a two dimension image.Based on this, the authors applied the artificial neurvl network edge detection algorithm to seismic section processing forthe prurpose of enhancing the quality of seismic section and, futhermore, the signal to noise ratio and resolution of seismic data.The application of algorithm to processing of the real seismic section has been proved to be effective and feasible.
出处
《物探与化探》
CAS
CSCD
2000年第5期353-357,共5页
Geophysical and Geochemical Exploration
关键词
人工神经网络
地震剖面
同相轴
分辨率
地震勘探
artificial neural network
image processing of seismic section
edge detection
phase axis
resolution
