AI Automatic First Break Picking Software
First-break picking is the basis and key link of seismic data processing, and plays an important role in the application of refraction wave static correction, shallow refraction exploration and tomographic imaging. First-break picking is a large part of seismic processing and is also the most tedious, especially in areas with complex near surface conditions and low signal-to-noise ratio, the difficulty and time cost of manual picking increases exponentially. Therefore, automated first-break picking has been an important topic for seismic processing. So far, there are many kinds of automatic picking methods, such as correlation method, fractal dimension method and neural network method, etc., but the application effect is often unstable, especially in picking data in complex area. At the same time, seismic acquisition technology advances bring explosive growth of seismic data, which makes the efficiency of massive data processing a new challenge.
On the basis of fully analyzing the requirements and characteristics of geophysical business, Wise-FBP first-break automatic picking system is designed and implemented with the Deep Learning Framework and Big Data technology. The system not only has the advantages of high processing efficiency of massive data, but also has high adaptability and reliable accuracy in complex area.
1. Massive data processing and high efficiency
With the advantages of Big Data Platform technology, especially for optimization of massive seismic data application, large-scale parallel computing based on GPU has been realized, which has big data processing ability and high efficiency.
2. Driven with AI technology, free manual intervention, strong anti-interference and good adaptability
Combined with traditional geophysical theory and information technology innovation, continuous training of a large number of different types of seismic data, ensures the algorithms more adaptive and higher accuracy.
3. Simple operation and flexible control
On the principle of minimizing manual intervention, the system takes full advantages of Deep Learning Framework, which is simple, flexible and easy to use.
Wise-FBP automatic first-break picking system is mainly centered on artificial intelligence parallel algorithms and developed together with various interactive quality control tools. At the same time, the system is based on the Big Data platform, which guarantees superior performance of massive data management and large-scale parallel computing. It also provides a variety of basic suits, such as project management, job editing management, auxiliary algorithm modules, interactive data quality control tools, and so on, as shown in Figure 1.3-1. These tools enable Wise-FBP system to have more comprehensive production and application capabilities, and smooth collaboration capabilities with other commercial software.