MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641093696 A) filed by S Singaravelan; Dr Balaji S; Dr. S. Usharani; Evangeline Nesa Priya D; Parameswari Manthiramoorthi; Deepthi S S; Bamila Rachel A; Agnes A; Sulakchana P; Mahadevi M; Elakkiyaselvam G; and Dr. Aravind B on August 03, 2026, for Foundation Model-Driven Remote Sensing Framework For Aircraft And Ship Analytics Using Multimodal Satellite Imagery.
Inventors include S Singaravelan; Dr Balaji S; Dr. S. Usharani; Evangeline Nesa Priya D; Parameswari Manthiramoorthi; Deepthi S S; Bamila Rachel A; Agnes A; Sulakchana P; Mahadevi M; Elakkiyaselvam G; and Dr. Aravind B.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: A foundation model-driven remote sensing framework for automated aircraft and ship analytics using multimodal satellite imagery is disclosed. The proposed framework integrates optical, synthetic aperture radar (SAR), infrared, hyperspectral, and multispectral satellite images to detect, classify, localize, monitor, and analyze aircraft and maritime vessels under diverse environmental and weather conditions. Initially, multimodal satellite images acquired from heterogeneous satellite platforms undergo preprocessing comprising radiometric correction, geometric correction, image registration, cloud removal, denoising, and spatial resolution enhancement. A pretrained vision foundation model extracts generalized semantic representations from each modality, while a multimodal fusion module combines complementary spatial, spectral, and contextual information. An adaptive object proposal network identifies candidate aircraft and ship regions, followed by a transformer-based recognition module that performs object classification, localization, and confidence estimation. A temporal analytics module tracks object movement across sequential satellite images for activity monitoring and anomaly detection. An explainable artificial intelligence module generates visual attention maps and confidence scores to improve decision transparency. The framework further includes an automated reporting engine that produces geospatial intelligence reports comprising object category, geographic coordinates, movement trajectory, confidence level, and detected anomalies. The disclosed invention improves detection accuracy, robustness against adverse imaging conditions, scalability across multiple satellite sensors, and operational efficiency for defense surveillance, maritime security, disaster response, border monitoring, environmental observation, and intelligent geospatial analytics.
Disclaimer: Curated by HT Syndication.
