An Dual-Threshold Method with Multi-Class Support Vector Machine for Oil Spill Detection in Ocean

  • Dr.J.Senthil Murugan,Dr.S.Sathya, Dr.Dr.Selvarani.P; Mrs.k.Mahalakshmi
Keywords: Oil Spill Detection, Image processing, Segmentation, Classification, Multi-Class SVM.

Abstract

Oil spill detection is one of the burgeoning research applied to save the public survive closer to coastel regions. This paper is mainly focused on public concern to make healthier in terms of coastel regions and environmental systems. Due to the spilled oil in water surface, water quality, acuastic animals, marine lives are getting damaged. In order to monitor the oil spill in all large area in all day under all weather  SAR and RADARSAT-2 SAR based sensors are used. The imaging mechanism comprised of speckle noises, patches and other problems like blur, rub on the image caused due to various nature and physical phenaomena which spoil the accuracy of oil spill detection. The Dual-Threshold method with Multi-Class Support Vector Machine is used to detect and segment the oil spills on the SAR images. This approach is implemented and results are verified in MATLAB software. The performance of the proposed approach is evaluated by comparing the results with the existing approach in term of Accuracy, Specificity and Sensitivity based on oil spill detection and classification. From the performance evalaution, it is proved that the proposed Dual-threshold method with MSVM endow with the finest results.

Published
2021-10-27
How to Cite
Mrs.k.Mahalakshmi, D. M. D. (2021). An Dual-Threshold Method with Multi-Class Support Vector Machine for Oil Spill Detection in Ocean. Design Engineering, 7318-7331. Retrieved from http://thedesignengineering.com/index.php/DE/article/view/5776
Section
Articles