MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202621041816 A) filed by Sampada P. Chaudhari; and Dr. Nileshsingh V. Thakur on April 01, 2026, for A Hybrid Colour Image Compression System Using Quadtree-Based Self-Similarity Analysis And Deep Learning Techniques.
Inventors include Sampada P. Chaudhari; and Dr. Nileshsingh V. Thakur.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: The present invention discloses a hybrid colour image compression system and method that integrates quadtree-based spatial decomposition, self-similarity analysis, and deep learning techniques to achieve efficient image compression with high reconstruction quality. The system receives an input colour image and processes it through a quadtree decomposition module that partitions the image into hierarchical blocks based on homogeneity criteria. A self-similarity analysis module identifies repetitive patterns within and across the blocks to eliminate redundant information. A deep learning-based compression module, comprising neural network architectures such as convolutional autoencoders, performs feature extraction, encoding, and optimization to generate a compact latent representation. The processed data is further encoded using a quadtree coding and matching mechanism to produce a compressed data stream. During decoding, the system reconstructs the image using the encoded structural and learned features, ensuring minimal distortion and improved perceptual quality. The proposed invention provides enhanced compression efficiency, reduced storage and bandwidth requirements, and improved adaptability across various image types, making it suitable for applications in multimedia systems, medical imaging, remote sensing, and real-time communication.
Disclaimer: Curated by HT Syndication.
