Date of Award
2024
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Computer Science
Committee Chair
Chaity Banerjee Mukherjee
Committee Member
Letha Etzkorn
Committee Member
Tathagata Mukherjee
Research Advisor
Chaity Banerjee Mukherjee
Subject(s)
Deep learning (Machine learning), Neural networks (Computer science), Land cover--Florida--Leon County--Remote sensing--Analysis
Abstract
This research delves into the intricate task of delineating land cover types in Tallahassee-Leon County and emphasizes the need for detailed granularity beyond existing classification systems. Utilizing cutting-edge GIS data, the study harnesses the power of deep learning algorithms, including U-net, UNetPlusPlus, FPNnet, and DeepLabV3Plus. A unique approach, ”Interlacing Unstructured Data with Deep Neural Nets,” integrates shapefiles and Tiff images to enhance classification metrics such as mean intersection over union, pixel accuracy, and loss functions. The research aspires to significantly improve the precision of land cover classification, holding implications for urban planning and environmental management. By innovatively integrating unstructured data, the study aims to offer valuable insights and tools for informed decision-making, contributing to urban development and environmental sustainability in Tallahassee-Leon County. The expected outcomes of this research carry profound implications for advancing our understanding of urban landscapes and fostering sustainable development practices.
Recommended Citation
Athmakur, Srivani, "Interlacing unstructured data with deep neural nets for predicting pervious and impervious land cover types" (2024). Theses. 671.
https://louis.uah.edu/uah-theses/671