MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202531129511 A) filed by Asansol Engineering College, Asansol, West Bengal, on Dec. 20, 2025, for 'adaptive multi-layer image restoration using context-aware neural filters.'
Inventor(s) include Dr. Apurba Chatterjee; Sumanta Karmakar; and Kaushik Neogi.
The application for the patent was published on April 17, under issue no. 16/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to an adaptive image restoration system that employs a multi-layer neural architecture with context-aware dynamic filtering to restore images degraded by noise, blur, compression artifacts, illumination imbalance, and structural distortions. The system incorporates a Context Extraction Unit that analyzes spatial, spectral, and semantic features of an input image to identify degradation characteristics. A Degradation Classifier determines the dominant distortion types, and a Dynamic Filter Selection Engine activates specialized neural filters accordingly. Restoration is performed through a Multi-Layer Restoration Pipeline consisting of coarse correction, structural enhancement, and fine-detail reconstruction stages. A Layer Fusion Module adaptively blends intermediate results using attention-driven weighting, while a Confidence Scoring Unit evaluates restoration quality and triggers refinement when necessary. The invention provides high-quality, computationally efficient restoration suitable for medical imaging, surveillance, satellite imaging, mobile photography, and real-time embedded vision systems."
Disclaimer: Curated by HT Syndication.
