MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641050643 A) filed by Mohan Babu University, Tirupati, Andhra Pradesh, on April 21, for 'spatiotemporal urban heat island estimation using a hybrid cnn-lstm deep learning framework with landsat 8/9 imagery.'
Inventor(s) include Mr. M. Venkata Naresh; and Dr. C. Venkata Sudhakar.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "Estimating UHI intensity is very important for urban resilience. While traditional pixel-based methods provide base-line data, they do not capture the complex spatial and temporal relationships in satellite images. This paper proposes a Hybrid Deep Learning Framework that integrates 1D-CNN, 2D-CNN, and LSTM architectures (models) for improved analysis of UHI. Landsat 8 & 9 data (2020-2021) will be used to extract Land Surface Temperature (LST), NDVI, and NDBI. We will provide a description of our application of Refined Lee Filtering (RLF) for pre-processing and Modified Binary Equilibrium (MBE) for feature selection as outlined in the cited references. Experimental results will show that the hybrid CNN-LSTM model will have an Overall Accuracy of and will significantly outperform stand-alone models. Our study demonstrates the importance of deep spatial-temporal features in mapping the growth of extreme thermal zones in highly urbanizing areas."
Disclaimer: Curated by HT Syndication.
