Intellectual Property India Publishes Patent Application for 'Scalable Deep Learning Framework For Efficient Satellite Image Segmentation And Classification' Filed by Deepa Raja; Dr. Suja A. Alex; Renees Jenisha J; Rathnaa Kumari M; and Deepa. R

MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202641043262 A) filed by Deepa Raja; Dr. Suja A. Alex; Renees Jenisha J; Rathnaa Kumari M; and Deepa. R, Chennai, Tamil Nadu, on April 4, for 'scalable deep learning framework for efficient satellite image segmentation and classification.'

Inventor(s) include Dr. Suja A. Alex; Renees Jenisha J; Rathnaa Kumari M; and Deepa. R.

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 describes an advanced edge-based system for satellite image segmentation and classification using a sequential deep learning architecture designed for real-time applications. The system combines two powerful models: U-Net, which performs precise pixel-level segmentation to identify different land cover regions such as buildings, roads, vegetation, and water bodies, and ResNet, which classifies these segmented regions with high accuracy by extracting deep features. Unlike traditional methods that rely heavily on cloud computing, this system processes data locally on edge devices, significantly reducing latency and bandwidth consumption. To enhance performance, the system incorporates preprocessing techniques such as image normalization, resizing, and noise removal, ensuring stable and efficient model training. Additionally, optimization methods like pruning and quantization are applied to reduce model size and computational complexity, making it suitable for resource-constrained devices like Raspberry Pi. The sequential approach ensures that segmentation identifies "where" objects are located, while classification determines "what" those objects are, improving overall efficiency. As a result, the system achieves high accuracy, faster inference, and scalability. This makes it highly suitable for applications in remote sensing, environmental monitoring, urban planning, and real-time geospatial analysis."

Disclaimer: Curated by HT Syndication.