• 应用地球物理学Ⅰ（油气及金属矿产地球物理勘探） •

### 基于非局部均值滤波的复杂地表区浮动基准面计算方法

1. 中国石油化工股份有限公司石油物探技术研究院,南京 210014
• 收稿日期:2018-03-28 修回日期:2018-08-20 出版日期:2018-10-20 发布日期:2019-01-11
• 作者简介:王胜春,男,工程师,1968年生,1997年毕业于长春科技大学,获大学学士学位,目前主要从事地震资料处理解释高性能支撑方面的研究.

### Method for calculating the floating datum in complex surface area based on NLM filter

WANG Sheng-chun(),LI Jin,HU Shan-shan

1. SINOPEC Geophysical Research Institute, Nanjing 210014, China
• Received:2018-03-28 Revised:2018-08-20 Online:2018-10-20 Published:2019-01-11

Abstract:

In the south and midwest regions of our country, since the complexity of the surface condition, when we use the mean static correction method to calculate the floating datum for prestack time migration, there will be sawtooth in the seismic section,which is due to the sudden change of static correction. Therefore, we need a smoother floating datum to seismic processing. At present, there are average static correction method and the smooth surface method to calculate floating datum statics correction. The datum obtained by different methods is different. The paper presents a method of calculating the smoothing floating datum statics by using Non-local mean filter, the floating datum statics at each CMP(Common Middle Point) point use Non-local mean filter algorithm to smooth in spatial domain, especially for 3-D seismic data. The basic principle of the non-local mean filter algorithm is similar to the mean filter algorithm, which requires the average value. But the Non-local mean filter adds the weight value of each point when calculating, so it can ensure that the influence between adjacent and large difference points is reduced when the average value is calculated in the frame. It also keeps a lot image edge details,so that the image looks clearer. In practical application, the first step is to calculate the statics of each shot point and detection point, then we can use the average statics method to calculate the average statics of each CMP point, at last we can use Non-local mean filter method to smooth the results of the second step. This method is more robust than the local mean filtering method, and can better protect the structural information of the signal, and the effect of the algorithm is better than other algorithms. The results show that the float datum statics calculated by this method is smoother and similar to the float datum statics calculated by the average static correction method. It can eliminate the sawtooth phenomena in the seismic profiles. The algorithm has a good effect, which proves the rationality and validity of the algorithm.