Underwater image enhancement and dehazing using wavelet based fusion for pipeline corrosion inspection

Abstract

Most of the underwater images contain a layer of haze, formed by suspended particles in the turbid water that create scattering and absorption of light. Absorption limits the visibility due to light attenuation while scattering blurs the image features, ultimately the underwater images captured by the camera in such a medium, are hazy and degraded as compared to the normal images taken in the atmosphere. In this paper, we are proposing a wavelet-based fusion method to enhance and dehaze the underwater hazy images of corroded pipeline. The main aim of this research is to inspect the underwater pipelines for the corrosion estimation. There are three main stages in the proposed method. At the initial stage, the hazy underwater image of corroded pipeline is enhanced by adjusting its contrast and the color profiles. In the second stage, the wavelet-based decomposition and inverse composition are performed to fuse the enhanced versions into a dehazed image. At the final stage, the corrosion on the surface of the pipeline is estimated. In order to validate the performance, the corrosion is estimated in both hazy and dehazed image.