Detection and Identification of Neighboring Pittings Using High-Resolution Optical Method

Abstract

Neighboring pittings are one of the most frequently found defects in gas pipelines. Due to the interaction with each other, these defects accelerate the corrosion growth rates in the wall of the pipelines. So, it’s meaningful to discriminate multiple neighboring pittings from single pittings by analyzing their properties. However, accurate characterization of these defects is a challenging task because of the limited resolution of the current inspection methods. This paper presents a high-resolution optical inspection method and automated sizing algorithms for estimating the diameters and separation distances of the neighboring pittings. The proposed optical method consists of a three-axis scanner, a linear laser diode, a CCD camera, and a test specimen containing various patterns of neighboring pittings. C-scan images for the defined defect geometries are generated by extracting the intensity information of the reflected laser light captured using a CCD camera. Neighboring pittings can be differentiated and characterized using the obtained C-scan images. Experiments in a lab configuration have been conducted and satisfactory results have been achieved.