MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202641064480 A) filed by Dhanavelu Anandan; Mrs. R. Suganya; Mr. R. Muthuselvan; Mr. S. Niranjan; Mr. Santosh. R; Mr. Kamesh. R; and Mr. Nivas. R, Karur, Tamil Nadu, on May 21, for 'multi modal deep learning framework for early forest fire detection using sentinel-2.'

Inventor(s) include Dhanavelu Anandan; Mrs. R. Suganya; Mr. R. Muthuselvan; Mr. S. Niranjan; Mr. Santosh. R; Mr. Kamesh. R; and Mr. Nivas. R.

The application for the patent was published on May 29, under issue no. 22/2026.

According to the abstract released by the Intellectual Property India: "Forest fires affect the degradation of the ecological environment, loss of economy and danger to the human life, especially in overpopulated and climatic sensitive areas. The effective mitigation and disaster management would therefore require the timely detection of fire outbreaks. The paper is a multimodal deep learning architecture of early forest fire detection based on multispectral imagery obtained by Sentinel-2. The suggested method is a combination of the visible (RGB), Near-Infrared (NIR), and Short-Wave Infrared (SWIR) to have both spatial patterns and spectral properties related to active fires and vegetation stress. Besides raw spectral bands, fire sensitive indices like Normalized Difference Vegetation Index (NDVI) and Normalized Burn Ratio (NBR) are also added to strengthen the discrimination between the burned areas and high risk areas. The proposed Convolutional Neural Network (CNN) with a feature-level fusion is created to simultaneously construct the complementary representations of multiple modalities to increase the reliability of detection in different atmospheric and seasonal contexts. The results of experimental assessment in a variety of forest landscapes prove to be more accurate, have fewer false alarms, and more robust than single-modality models. The suggested model will provide a scalable and automated satellite-based early warning system, which can be used to maintain proactive forest management and environmental protection strategies."

Disclaimer: Curated by HT Syndication.