Ship Pipeline Corrosion Surface Evaluation Using Image Processing in MATLAB

Image Processing Toolbox is one of the apps in MATLAB which allow a user to automate standard image processing processes. The user can segment image data in real-time, benchmark image registration methods, and batch process large datasets. In shipyard industry, the image processing technique is a potential method to do corrosion inspection instead of visual inspection to achieve reliable method and give quick results. This paper investigates a corrosion method specifically surface area. The corrosion image with different areas is improved by using image enhancement and noise removal technique. All images are then evaluated using image analysis whereby extracting useful information from images, such as shape recognition, object tracking, coloring, or object characteristics measurement. For quality results, MSE (Mean Square Error) and PSNR (Peak Signal Noise Ratio) are applied. The results demonstrate that image processing effectively calculates the corroded area.