Artificial Neural Network and Gray Level Co-Occurrence Matrix Based Automated Corrosion Detection
V.Thanikachalam1, M.G.Kavitha2, V.Bharathi3
1Dr.V. Thanikachalam*, Associate Professor Department of IT , SSN College of Engineering, Chennai.
2Dr. M. G. Kavitha, Assistant Professor, Department of CSE, University College of Engineering, Paddukottai.
3V. Bharathi, Assistant Professor, Department of ECE, Kongunadu College of Engineering and Technology, Thottiyam, Trichy.
Manuscript received on July 20, 2019. | Revised Manuscript received on August 10, 2019. | Manuscript published on August 30, 2019. | PP: 4499-4502 | Volume-8 Issue-6, August 2019. | Retrieval Number: F9000088619/2019©BEIESP | DOI: 10.35940/ijeat.F9000.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this paper, we show an image processing algorithm with its capabilities in detecting the corrosion. This algorithm is programmed and requires no parameter modification and no previous knowledge of image acquisition process because function evaluates their parameters. Digital image processing technique proposed to avoid such incident occurrences. Combining Poisson-Gaussian- Mixture distribution with a Fuzzy segmentation framework an algorithm is developed to clutch image information. Artificial neural network and gray level co-occurrence matrix (GLCM) utilized to recognize the corrosion. The developed algorithm can be used in the ROV to detect the corrosion spots. The algorithm results exhibit the sufficiency in perceives corroded spots. Using image processing the corrosion detection process can be automated with a monitoring software setup which can generate an alert based on corrosion severity. Using image processing the infrastructure’s corrosion evaluation effort will be minimized, and presenting the result statistics is easier. In application point of view, we can extend the algorithm capabilities to the fatigue crack detection.
Keywords: Image Processing, Corrosion Detection, Image denoising.