A STEEL BRIDGE RUST DISCRIMINATION APPROACH COMBINING SUPPORT VECTOR MACHINE AND NEURAL NETWORKS

摘要

One of the traditional methods to assess the quality of the coating on the steel bridge surface is calculating the ratio of rust to the total area of coated surface. In North America, digital image processing has been applied on quality assessment of steel bridge coating since late 1990s due to the difficulty of rust area quantification by human vision. Although a brand-new method, Support-Vector-Machine-based Rust Assessment approach, was proposed to differentiate defective images from non-defective images efficiently, it only simply segmented images into two groups, the rust and the background, like most of the previously developed systems, without recognizing the intensities of rust defects. A new steel bridge rust intensity assessment method which hybrids support vector machine and Artificial Neural Networks is proposed in this paper for automatically recognizing rust defects and identifying their intensity to describe gradually changed colors.