A New Method for High Resolution Seismic Coherency Attributes Computing

In order to improve the effects of interpretation of small faults and other geo-bodies based on the seismic data, we proposed a new method to increase the resolution of current coherency attributes analysis. The proposed method is mainly based on the principle of medical image enhancement technology, which firstly do the convolution of Gauss Kernel function with C3 coherency data, then carry out the two order partial derivative of the convoluted data to get maximum eigenvalue as the high resolution coherency results. The numerical and field data examples have proved that the resolution of the new method is increased obviously, especially with high lateral resolution. The results also show more clearly the spatial distribution of faults, and the details of the subsurface structure which are very important for small faults interpretation.