Rock Solid Images

Seismic Facies Modeling

Designed to complement ATTRIB3D®, LithANN® uses advanced neural network algorithms to define regions of common attribute response or seismic facies.

LithANN® allows the user to combine two or more seismic attribute volumes to maximize the discriminatory capabilities of each attribute. LithANN® integrates seamlessly with ATTRIB3D®, but attributes from third-party applications can easily be incorporated.

LithANN® offers several classification algorithms, including feed-forward back-propagating artificial neural networks (ANN) and Kohonen Self Organizing Map (KSOM) methods. Also incorporated are a variety of diagnostic tools to aid in attribute selection and parameterization.

The output from LithANN® is a classified volume where each seismic sample is replaced by a class value representing the seismic facies unit. The LithANN® output is interpreted within a voxel-based 3D volume based visualization system; Rock Solid Images has developed a number of specialized interpretation workflows designed to maximize the benefit of seismic facies modeling combined with 3D visualization.