About

GeoMatics is a GIS research group at the UCSC dedicated to pushing the boundaries of geospatial technologies by addressing current limitations and developing innovative solutions across various domains. Our objectives encompass the development and implementation of cutting-edge methodologies for spatial analysis in areas like urban planning, environmental science, public health, disaster management, and natural resource management. We advocate for the widespread use of GIS in decisionmaking processes, emphasizing the importance of open source software and data accessibility for research and development. Collaborative interdisciplinary efforts are encouraged, fostering partnerships between professionals and students from diverse fields such as geography, computer science, environmental science, public health, and engineering. Through knowledge sharing, capacity building, and a supportive research environment, we aim to sustain a continuum of GIS-related explorations and contribute to dissemination of impactful findings through publications, workshops, and conferences. Some of the research areas that the group focuses on include:
  • Spatial analysis and modeling
  • Remote sensing and GIS integration
  • Spatial data visualization
  • GIS applications in social sciences
  • GIS for disaster management and risk assessment
  • Geospatial data analysis for transportation and logistics
  • GIS for environmental sciences, including wildlife management and conservation

Projects

Completed Projects

  • Estimation of Land Coverage Utilized for Pepper Plantations in Sri Lanka through Drone Imagery Analysis
  • Detection of Marine Plastic in Sri Lanka Utilizing Multispectral UAV Images
  • Uncertainty Visualisation and Real-time Detection of Wild Elephant Locations

Research Team

Principal Investigator

Research Team

Research Assistants

  • Ms. S. A. V. S. Piyathilake – (from 1st June 2023 – 31st December 2023)

Publications

Journal Article

  • Dissanayake, V. Piyathilake, A. P. Sayakkara, E. Hettiarachchi, and I. Perera, “Eloc-Web: Uncertainty Visualization and Real-Time Detection of Wild Elephant Locations,” Journal of Geovisualization and Spatial Analysis, vol. 8, no. 1, p. 7, Jan. 2024, doi: 10.1007/s41651-023-00169-7.

Available: https://doi.org/10.1007/s41651-023-00169-7.

Impact factor: 4.0 (2022)

Work Summary: This paper discusses the urgent need for proactive conservation measures to address the substantial decline in elephant populations, primarily driven by human-elephant conflict. Conservationists are increasingly adopting passive acoustic monitoring for tracking wild elephants, but the time-consuming data analysis poses a challenge for real-time tracking. In response, the study introduces Eloc-Web, a web application that leverages machine learning to visualize real-time elephant
locations based on classified acoustic recordings. The application considers uncertainties in audio classification, ensuring accurate representation. Developed through a user-centered design process with input from elephant ecologists, Eloc-Web demonstrates promising potential in facilitating real-time data visualization for the conservation of elephant populations.

Research Domains: GIS (Uncertainty Visualization, Geovisualization), User-Centered Design

Conference Proceedings

  • Purushoth, V. Piyathilake, K. Athapaththu, D. Sandaruwan, A. P. Sayakkara, and E. Hettiarachchi, “Using Multispectral UAV Imagery for Marine Debris Detection in Sri Lanka,” International Conference on Information Technology Research (ICITR 2023), University of Moratuwa, Sri Lanka: IEEE Xplore, Dec. 2023.

Available: https://doi.org/10.1109/ICITR61062.2023.10382866

Work Summary:  This study addresses the issue of marine pollution in Sri Lanka, emphasizing the country’s significant contribution to marine debris. The impact of marine pollution on biodiversity, fishing, and tourism is highlighted. Traditional monitoring methods involve labor-intensive processes, but there is a growing trend worldwide, including Sri Lanka, to use Unmanned Aerial Vehicle (UAV) imagery for its efficiency and cost-effectiveness. The study focuses on using multispectral UAV imagery to monitor marine debris in Ambalangoda, Sri Lanka. We employed the CutPaste method for region detection and the ResNet-18 model with Faster R-CNN for classifying marine debris instances in the images. The implemented approach achieved a classification accuracy of around 60%, showcasing potential for future improvements in automated marine debris detection.

Research Domains: GIS (Remote Sensing), Image Processing, Machine Learning