Underwater Image Identification Using Fuzzy Soft Planar Graph

Authors

• Bhuvaneswari Natarajan Jayakar, Department of Mathematics School of Advanced Sciences KalasalingamAcademy of Research and Education (Deemed to be University) Krishnankoil – 626 126, Tamil Nadu, India
• Karthick Palanisamy, Department of Mathematics School of Advanced Sciences Kalasalingam Academy of Research and Education (Deemed to be University) Krishnankoil – 626 126, Tamil Nadu, IndiaFollow

Corresponding Author

Karthick Palanisamy

Subject Area

Computer Science

Article Type

Special Issue Original Study

Abstract

High-attribute underwater image segmentation is fundamental to autonomous marine exploration, biodiversity monitoring, and subsea infrastructure inspection. However, underwater environments are inherently stochastic, exhibiting severe light attenuation, chromatic distortion, scattering effects, and noise, all of which significantly degrade image quality and challenge conventional computer vision techniques. Hence, Fuzzy logic which handles uncertainty and imprecision in data and Soft sets which manage parameterized information, are widely applied to underwater image analysis. Fuzzy Soft Planar Graphs (FSPGs) constitute a mathematical framework that integrates fuzzy set theory, soft set theory and planar graph structures for preserving spatial topology. The proposed methodology adopts a hybrid processing pipeline that begins with underwater-aware superpixel generation, enabling local feature aggregation and reduced computational complexity while maintaining fine structural details. Subsequently, region-level fuzzy soft modelling is employed to facilitate robust classification of complex marine entities, including fish, coral formations, and geological substrates. By incorporating hesitation degrees alongside spatial constraints, the framework effectively mitigates ambiguity in low-visibility underwater imagery. When the identification step operates on the contracted FSPG, noise is reduced the computational complexity decreases. Experimental evaluation on an underwater datasets demonstrates improved boundary fidelity and mean Intersection-over-Union (mIoU), compared with conventional graph- based and CNN-based baselines.

Keywords

Marine Object Detection, Fuzzy Soft Planar Graph, Graph Contraction, Superpixel Segmentation, Underwater Image Segmentation, Uncertainty Modelling

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 License.

Recommended Citation

Natarajan Jayakar, Bhuvaneswari and Palanisamy, Karthick (2026) "Underwater Image Identification Using Fuzzy Soft Planar Graph," Mansoura Engineering Journal: Vol. 51 : Iss. 4 , Article 21.
Available at: https://doi.org/10.58491/2735-4202.3491