Anomaly detection and visualization using color-stack, cross-plot, and anomalousness volumes
Tracy J. Stark · 2005
The search for hydrocarbons can be viewed as an anomaly detection and isolation problem. In general the largest anomalies have already been found; the large bright spots, flat spots and structures. Tools and methodology to find the smaller anomalies are now required. This presentation discusses a general technique of using cross-plot, anomalousness and color-stack volumes in combination with visualization software to help detect, and isolate anomalies that might otherwise “slip through the cracks”. The visualization software performs real-time volume sculpting by modifying the opacity lookup table of the anomalousness volume (which is used as a classification volume). The techniques will be illustrated using partial and full stack data over a 1000 ms window from a marine 3D data set. These partial and full stack volumes will be used to create the color-stack, cross-plot, and anomalousness volumes, which in turn will be used in visualization software for the anomaly detection. The methods discussed are not limited to the type of data used in this presentation, but with this type of data, more than one class of AVO anomaly is detectable.