Adaptive Fuzzy Clustering for Efficient Segmentation of Wellbore Resistivity Images

Abstract

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractResistivity image segmentation allows mapping of similar regions in a scene leading to the recognition of distinct geological features such as rock fractures and thin beds. Recognition of similar patterns embedded in resistivity image data is the basis of clustering regions in an image. Therefore an efficient clustering technique is a natural choice for image segmentation. Efforts to develop algorithms for adaptive and less computationally complex classification of data have led to the implementation of neural network classifiers of both supervised and unsupervised learning. Such neural network architectures are ways to achieve autonomous processing of patterns but are not considered to incorporate intelligent decision processes offered by various techniques of fuzzy clustering. Integration of fuzzy membership values of objects into neural network processing generates more robust models for intelligent pattern recognition.This paper presents and efficient object extraction method for resistivity image segmentation based on neuro-fuzzy clustering algorithms. The neuro-fuzzy approach retains the basic properties and architectures of neural networks and fuzzifies some other elements, enhancing the learning speed and generalization capabilities.