CORROSION DAMAGE ANALYSIS USING IMAGE PROCESSING

The corrosion of the pipelines or storage tanks are the potential failures for the power plant operation causing in the low efficiency and costly maintenance. Hence, the effect of corrosion has become a common problem to the industry. The current method applied is by visual inspection of identification and detection if there is atmospheric corrosion damage on metallic surfaces. This method is very tedious and subjective due to different level of expertise and experience. There is a need to develop a low cost automatic corrosion damage identification. This project focuses on early detection of pipelines and gas tanks corrosion in oil and gas. This is to avoid any unwanted incidents in the near future. This project will analyze the texture feature extraction techniques, select and apply the most appropriate technique to corrosion detection problem. The corrosion pattern is to be classified by applying neural network algorithm. A low-cost computational non-destructive evaluation (NDE) technique will be used to analyze the reliability of a system or component without interrupting its performance and operation. NDE is an automatic detection method for corrosion classification which can be implemented for the inspection of the pipelines and storage tanks in the oil and gas industry as it will not disturb their properties and performance.