Horizon And Fault Detection Research In 3D Seismic Base On Image Processing

In seismic exploration, seismic data interpretation is a very important part, and horizon and fault interpretation is of one of the best core part. Traditionally, three dimensional seismic interpretation methods are artificial, usually got digital seismic traces through the profile to form the horizon line, and then interpolated surface layer, not only repeated the results can not be verified, but also require a lot of manpower and time. Therefore, the paper around the three-dimensional seismic horizon and fault identification and automatic extraction, based on image processing carried out 3D seismic horizon and fault automatic interpretation algorithm, and designed the prototype system.First of all, the seismic trace-based approach prone to error or searches the wrong layer when encountered fault or weak signal areas, based on a strong orientation, we use watershed filtering algorithm and orientation-filtering enhancement the seismic image, make the layer axis signal much more continuous. In the tracking process, according to the layer's enhanced results and orientation to seed dispersal searching, simultaneously combined with the amplitude of seismic-based matching method to adjust search results, make the results in the seismic wave crests. Experimental results show that the algorithm can be adapted to a better signal to fulfill automatic distraction.Then, based on mostly current methods can only handle big fault with obviously linear features, combined with the thought of manual interpretation of two dimensional seismic profile, this paper proposed a human interaction and automatic identification jointly three-dimensional seismic fault interpretation method. Structural-oriented filtering and trace semblance were used to enhance the fault, and use the two-dimensional layer's constraints and corner detection to extract fault points, and then using image recognition to extract fault lines. By test and analysis, the accuracy of this method breakpoint detection is about 20%, can help explain the whole staff of geological structure, provide a reference for the search for breakpoints.Finally, developed a PC-based prototype system of automatic interpretation based on the proposed automatic interpretation algorithm, and fulfill the visual result of the algorithm.