Paper Accepted in IEEE IGARSS 2024

Paper published in IEEE InGARSS 2024

Journal Paper under Under Review (Revised Submission) in Trees, Forests and People (Elsevier)

Led a team in developing a novel tool that integrates machine learning (YOLOv8) and image processing techniques to improve tree detection accuracy and provide detailed health insights

Integrated Digital Elevation Models (DEM) and Digital Terrain Models (DTM) with image processing techniques, leveraging slope-based analysis to enhance tree detection and mitigate errors caused by dense undergrowth and shrubs to prevent false detections

Implemented a custom region-growing algorithm that increased tree detection accuracy from 40% to 85% in dense regions and 95% in sparse regions, while accurately measuring tree height, canopy girth, and health parameters

Built an interactive WebGIS platform using HTML, CSS, JavaScript, Flask, PostgreSQL, and GeoServer, allowing users to visualize data and access detailed information on urban forests for informed decision-making

Applied the methodology to a diverse range of tree species in urban environments, showcasing its versatility across different species and canopy shapes, unlike traditional approaches focused on specific trees