Hyperspectral Vision Based Probe for In Situ Corrosion Monitoring in Saline Environments

Abstract

Corrosion is a major concern in various engineering fields. Early detection of corrosion helps in reducing the risk associated with the failure of metal structures invaded by corrosion. Though several techniques exist for monitoring corrosion, most of these techniques are contact-based and tend to be destructive in nature. In addition, only a few of these techniques can provide fast, point-by-point analysis of corrosion. In this paper, we propose and demonstrate a novel corrosion inspection probe with integrated illumination for the sensitive monitoring of corrosion along each point of an area under inspection based on snapshot hyperspectral vision. Being a small and flexible probe, it can be used for monitoring corrosion in difficult-to-access locations. The experimental results correlate well with the standard corrosion prediction model and are also validated by spectroscopic Raman analysis. The possibility of incorporating automated corrosion monitoring based on machine vision as the future development of the probe is also highlighted.