Automated Detection of Corrosion Damage in Power Transmission Lattice Towers Using Image Processing

Corrosion is a serious issue causing damage in power transmission lattice towers of steel that can lead to outages. In spite of initial galvanization, periodic repainting and usage of weathering steels, corrosion is experienced in lattice structures at locations with constant exposure to moisture and inaccessibility to repaint. In the present day there are a huge number of transmission towers built of carbon and weathering steels with corrosion posing serious threats to their bearing capacity and durability. Timely inspection and repair is essential to avoid unprecedented structural failures. Employing non-destructive methods of manual inspection for large number of towers to detect corrosion and related damages is time consuming and expensive. In addition to this, the drawbacks include error due to inaccurate human judgment questionable safety of inspector to climb structures possibly weakened by corrosion. In such a situation non-contact approach of automated visual inspection for corrosion and related damage detection through image processing of aerial or ground based images is a viable option. A combination unsupervised and supervised classification methods is used to identify corroded regions in power transmission tower images using various color features of the image. The image is segmented into clusters based on the colors in L*a*b* color space using K-means clustering algorithm. These segments are tested against the conditions of hue obtained from statistical analysis of hue values corresponding to a set images of corroded surfaces to identify the segment of the image with corrosion. This approach lays a foundation for content based image retrieval in the domain of corrosion detection that is the ability to identify corroded structures from a large database of inspection images.