Monitoring system for corrosion in metal structures using a probe based hyperspectral imager

Corrosion in metal structures is one of the prevailing problems impacting automobile, cargo, and construction industries. The detection of corrosion at the right time and determination of the root cause are crucial in its prevention and control. In this context, we propose hyperspectral imaging as a potential imaging modality for monitoring corrosion. This technique is very relevant for high-speed, non-destructive inspection. The proposed hyperspectral imager can efficiently monitor corrosion with high sensitivity and it enables corrosion detection even at human inaccessible areas with the aid of a custom fabricated fiber optic probe. In contrast to traditional methods, the hyperspectral imaging technique can capture reflectance at several wavelengths from several spatial points of the sample and hence provides a means of rigorous analysis of the sample reflectance. Using a two dimensional to one dimensional fiber bundle reformatter, hyperspectral images of metal samples were recorded. Induced corrosion in the sample was monitored by the hyperspectral imager and the data recorded were processed to form the three-dimensional spectral datacube. Obtained results show that hyperspectral reflectance imaging is a powerful tool for corrosion monitoring, non-destructively.