Selective Segmentation of the Shallow Electrical Profiling Data

Abstract

A method for selective segmentation that correctly recovers the boundaries of objects in the plan and estimates the depth range is proposed. This method requires that the electrical exploration data are processed: shallow electrical profiling and vector images of the main directions of changes in the resistance of the medium. An object boundary is estimated on the basis of a fuzzy clustering algorithm (fuzzy c-means), adapted for analyzing multispacing electrical profiling. The depth range of an object in soil is estimated on the basis of applying a scalar product function to vector images of the main directions, adjacent in depth. The effectiveness of this method is confirmed by computer simulation and an example of data processing of the field experiment on an archeological site, which is a medieval fortified settlement.