OICC Press Fuzzy Optimization and Modeling Journal (FOMJ) 2676-7007 7 1 2026 03 30 Contour-Based Video Inpainting using Neutrosophic Sets 10.57647/fomj.2026.0701.03 EN Amanna Ghanbari Talouki Department of Technical and Engineering, Ayandegan University, Tonekabon, Iran https://orcid.org/0000-0001-5326-4075 Abbas Koochari Department of Computer Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran https://orcid.org/0000-0003-0584-6470 S. Ahmad Edalatpanah Department of Applied Mathematics, Ayandegan University, Tonekabon, Iran https://orcid.org/0000-0001-9349-5695 Journal Article 2026 03 30 Image and video inpainting represent critical challenges within the domain of visual information processing, serving pivotal roles in tasks such as restoration, reconstruction, and occlusion removal. The fundamental objective in these applications is to seamlessly restore missing or corrupted regions without introducing perceptible artifacts or structural inconsistencies. While image inpainting operates within a spatially static framework, video inpainting must additionally contend with temporal continuity, necessitating rigorous preservation of motion coherence and inter-frame consistency. This study presents an advanced inpainting methodology grounded in Neutrosophic logic, a mathematical framework adept at modeling uncertainty and indeterminacy within both spatial and intensity domains. By leveraging Neutrosophic-based segmentation, the proposed approach effectively isolates uncertain regions and guides the subsequent restoration process with enhanced precision. The primary objective of this research is to propose a novel approach for constructing an optimal contour database, which serves as a foundation for achieving superior foreground completion. This method aims to enhance the accuracy and visual coherence of the inpainting process by leveraging precise contour information, ultimately improving the overall quality of video restoration. Comprehensive experimental evaluations validate the efficacy of the proposed method, demonstrating notable improvements in perceptual quality and structural fidelity when benchmarked against state-of-the-art alternatives. Nonetheless, certain limitations persist, particularly in processing scenes characterized by high semantic complexity and rapidly changing backgrounds, which constitute promising directions for future investigation.