MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641094676 A) filed by Dhanalakshmi Srinivasan College Of Engineering And Technology on August 05, 2026, for Deep Learning Enhanced Hybrid Demosaicking With Explainable Hvs Modules.

Inventors include Dr K John Peter; Avanthika U; Gowthami D; and Monikasri S.

The application for the patent was published on September 11, 2026, under issue no. 37/2026.

Abstract: Abstract of the Invention Deep learning enhanced hybrid demosaicking with explainable HVS modules Image demosaicking is a critical image processing technique used in digital cameras to reconstruct full-colour images from raw sensor data captured through a Colour Filter Array (CFA), commonly the Bayer pattern. Since each pixel in the sensor captures only one colour component red, green, or blue the missing colour information must be estimated to generate a complete image.Accurate demosaicking is important because it directly influences image quality, colour accuracy, and visual clarity in digital imaging applications.Traditional demosaicking methods mainly rely on interpolation algorithms to estimate missing pixel values. While these approaches are simple and computationally efficient, they often produce visual artifacts such as colour distortion, zipper effects, false colours, and blurring, especially near edges and textured regions. These limitations reduce the quality of reconstructed images and create the need for more advanced and intelligent reconstruction techniques.To overcome these issues, this project proposes "Deep Learning Enhanced Hybrid Demosaicking with Explainable HVS Modules." The proposed system combines conventional interpolation methods with deep learning-based enhancement for improved reconstruction performance. Initially, the raw CFA image undergoes preprocessing steps such as normalization and noise reduction to improve input quality. The system then applies hybrid demosaicking and deep neural network refinement

Disclaimer: Curated by HT Syndication.