MUMBAI, India, June 22 -- Intellectual Property India has published a patent application (202641070086 A) filed by Eswar College Of Enginering Autonomous on June 04, 2026, for Semantic Segmentation Of Seismic Images Using Deep Learning Techniques.
Inventors include Dr. Bolla Ramesh Babu; and Dr Shaik Muzeer.
The application for the patent was published on June 12, 2026, under issue no. 24/2026.
Abstract: This invention introduces an intelligent Semantic Segmentation System for Seismic Images powered by advanced Deep Learning Techniques, designed to accurately identify and delineate subsurface geological structures from complex seismic data in a fully automated and non-invasive manner. The proposed system processes raw seismic images through a specialized convolutional neural network architecture enhanced with attention mechanisms and multi-scale feature fusion, enabling precise pixel-level classification of various subsurface features such as faults, horizons, salt bodies, and stratigraphic layers. By integrating transfer learning from pre-trained models with custom seismic-specific data augmentation, the system achieves superior boundary detection and reduces interpretation uncertainties that commonly affect traditional manual or semi-automated seismic analysis methods. Real-time processing capabilities combined with uncertainty estimation modules make this invention highly reliable for field applications, offering enhanced accuracy, faster decision-making, and significant cost savings in hydrocarbon exploration and geophysical studies.
Disclaimer: Curated by HT Syndication.
