Noise-Produced Patterns in Images Constructed from Magnetic Flux Leakage Data

TL;DR

This paper analyzes noise-induced patterns in magnetic flux leakage images used for detecting corrosion in steel casings, focusing on their scaling properties and implications for defect resolution.

Contribution

It introduces an analytical approach to evaluate noise patterns in flux leakage images, applicable to complex linear transforms beyond direct measurements.

Findings

Noise patterns exhibit specific scaling behaviors.

Analytical methods enable better understanding of noise impact.

Approach can be adapted to other imaging techniques.

Abstract

Magnetic flux leakage measurements help identify the position, size and shape of corrosion-related defects in steel casings used to protect boreholes drilled into oil and gas reservoirs. Images constructed from magnetic flux leakage data contain patterns related to noise inherent in the method. We investigate the patterns and their scaling properties for the case of delta-correlated input noise, and consider the implications for the method's ability to resolve defects. The analytical evaluation of the noise-produced patterns is made possible by model reduction facilitated by large-scale approximation. With appropriate modification, the approach can be employed to analyze noise-produced patterns in other situations where the data of interest are not measured directly, but are related to the measured data by a complex linear transform involving integrations with respect to spatial coordinates.