MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202541008202 A) filed by K. S. Rangasamy College Of Technology on January 31, 2025, for Vlsi Implementation Of Random Noise Removal In Digital Images Using Hybrid U-N.
Inventors include Mr. S. Pradeep; Mr. S. Saravanan; Mr. E. Chandrakumar; Mr. Mohanapraveen T; Mr. Monish G; and Mr. Sathyan R.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: Random noise removal in digital images is a critical challenge in image processing, particularly for applications requiring high-quality, noise-free images such as medical imaging, satellite data, and security systems. This work proposes a hybrid technique for random noise removal using the Hybrid U-NET architecture, designed for real-time image denoising. The proposed method integrates advanced deep learning models with VLSI hardware acceleration to achieve efficient and high1quality noise removal. The Hybrid U-NET combines convolutional neural networks (CNNs) with skip connections and hybrid layers, which allows for the suppression of noise while preserving important image details. The system operates in two stages: noise detection using the U-NET model and noise removal using a fusion of learned features. This dual-stage approach enhances the precision and effectiveness of the denoising process, ensuring high Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM) for the restored images. The VLSI-based implementation accelerates the computation process, enabling real-time performance even in resource-constrained environments such as mobile devices and embedded systems. Additionally, the integration of deep learning techniques ensures that the system adapts to different types of noise and images, providing optimal denoising performance across various applications.
Disclaimer: Curated by HT Syndication.
