Automated Box Data Matching for Multi-Modal Magnetic Flux Leakage Inspection of Pipelines

Abstract

Magnetic flux leakage (MFL) is one of the most popular in-line inspection techniques to detect pipeline corrosion defects. However, the measurement limitation and random error associated with the individual MFL could introduce observable variations to the inspection. One potential way to reduce such variation is to fuse the inspection data from axial MFL (aMFL) and circumferential MFL (cMFL) due to their complementary measurement nature. Data from the multi-modal MFL must be matched before the fusion operation. However, there is no literature reporting the multi-modal MFL data matching yet. In this article, we propose an automatic box data matching framework that consists of four major steps, i.e., pre-processing, grouping, alignment, and matching. The pre-processing step accomplishes the 3-D to 2-D data mapping and cleansing. The grouping of the box data from the cMFL based on their location information is given in the following. Then, the coordinate systems of two MFL data sets are aligned. Finally, in the aligned coordinate system, the density-based spatial clustering of applications with noise (DBSCAN) algorithm is revised to perform many-to-many matching. The experimental results demonstrate that the proposed framework can accurately match the multi-modal MFL box data. Thus, this solution will enable the multi-modal MFL data fusion for a more reliable and accurate pipeline corrosion assessment.