Restoring and Enhancing Degraded Underwater Pipelines for Identifying and Detecting Corrosion

Abstract

Detection of corrosion from underwater images is necessary for oil and gas pipelines to eliminate the internal leakages and hazards. The tests utilized a broad range of underwater pictures of various situations. A modern technique for estimating subsea pipeline corrosion based on the colour of the corroded pipe. For corrupted underwater videos, an image reconstruction and enhancement algorithm is created as a preliminary phase. The created algorithm reduces blurring and improves picture colour and contrast. The improved colours in the imaging details aid in the method of corrosion estimation. In this work we proposed a underwater corrosion detection using image processing techniques. Some machine learning and deep learning techniques have been used for classification of corrosion. In experimental analysis various features have been evaluated for detection of corrosion and it introduces better classification accuracy than traditional approaches.