
Full text loading...
Various methods of averaging or smoothing data fields are commonly used to suppress random noise and enhance anomaly characteristics for interpretation. A new approach, termed smart averaging, which reduces uncertainty in determination of anomaly size and location is introduced in this article. In this procedure, computation of each output value is considered to be an optimization problem based upon some criterion, or set of criteria, with different benefits for interpretation (e.g., sharp anomaly delineation, smoothing without suppressing anomalies, or reducing random noise or noise peaks without affecting anomaly characteristics). A subset of averaging window parameters (location, shape, and size) is tested for each grid point before some optimal window parameters are found, based upon some chosen criterion. Smart averaging can be useful in various data processing applications for interpretation purposes, such as data interpolation, image resolution and analysis, and correlation of well logs.