Page 38 - AJWEP-v22i3

Asian Journal of Water, Environment and Pollution. Vol. 22, No. 3 (2025), pp. 32-45.
doi: 10.36922/ajwep.8282

ORIGINAL RESEARCH ARTICLE

SpillNet: A modified convolutional neural network model
for oil spill detection

Tokula I. Umaha 1,2 , Felix Ale 1,2 , Ikpaya D. Ikpaya 1,2 , John A. Momoh 1,2 ,
Steve A. Adeshina , Ilesanmi A. Daniyan * , and Adeyinka P. Adedigba 5
4
3
1 Department of Systems Engineering, School of Engineering, African University of Science and Technology, Garki, Abuja,
Nigeria
2 National Space Research and Development Agency, Institute of Space Science and Engineering, Abuja Obasanjo Space
Centre, Abuja, Nigeria
3 Department of Electrical and Electronics Engineering, Faculty of Engineering, Nile University of Nigeria, Abuja, Nigeria
4 Department of Mechatronics Engineering, College of Engineering, Bells University of Technology, Ota, Ogun, Nigeria
5 Department of Mechatronics Engineering, Faculty of Engineering and Technology, Federal University of Technology,
Minna, Niger, Nigeria
*Corresponding author: Ilesanmi A. Daniyan (iadaniyan@bellsuniversity.edu.ng)

Received: December 29, 2024; Revised: February 10, 2025; Accepted: February 21, 2025; Published Online: March 6, 2025

Abstract: Rapid and accurate detection of oil spills is crucial for initiating timely response measures to mitigate
environmental impacts. This study proposes an oil spill detection method based on a modified convolutional
neural network, termed “SpillNet.” The architecture integrates multiple depthwise separable convolutional layers,
batch normalization, and residual connections to enhance feature extraction and learning capabilities. The dataset
consists of synthetic aperture radar images obtained from Sentinel-1 satellites, part of the European Space Agency’s
Copernicus program. Model training was conducted on an NVIDIA Tesla T4 GPU available on Google Colab, with
up to 12GB of random access memory. Programming was carried out in the Python environment using Python
3.7, and all required libraries were installed through pip. The results indicate that the proposed model achieves
an accuracy of 0.946947, a mean Intersection over Union of 0.58124, and a mean specificity of 0.944469. These
results demonstrate that the proposed model outperforms existing models in the oil spill segmentation task. This
study contributes to advancing automated oil spill detection by offering a reliable and efficient solution for early oil
spill detection and environmental monitoring.

Keywords: Batch normalization; Convolutional neural network; Model; Oil spill; Programming; Synthetic aperture
radar

1. Introduction include oil tanker accidents, damaged ships or pipelines,
illegal bunkering, and the indiscriminate disposal of oil
Oil spills pose a significant threat to marine and into water bodies, among others. Rapid and accurate
1
coastal environments, causing long-lasting damage detection of oil spills is crucial for initiating timely
to ecosystems, wildlife, and human health. If not response measures to mitigate their impact, protect
controlled, they can result in environmental pollution marine life, and ensure a clean and safe environment.
that devastates marine life. Major sources of oil spills Over the years, methods, such as aerial photography

Volume 22 Issue 3 (2025) 32 doi: 10.36922/ajwep.8282