AI-Enabled Projects

Remote sensing, geospatial analytics, and embedded AI applications

A portfolio-style gallery for selected publications, deployed tools, and applied experiments. This page stays static and presentation-focused, while heavier interactive runtimes can live on dedicated app infrastructure.

Selected Projects

Forest qualitative experiment
LiDAR 3D segmentation Qualitative analysis

Qualitative Experiments: Cross-Ecosystem Transfer from Forest to Sparse Semi-Arid Rangeland Vegetation

This portfolio entry presents qualitative comparison experiments of instance-level and semantic segmentation outputs generated by SegmentAnyTree and ForestFormer3D. A live embedded viewer is provided within the same web application.

Focus: instance and semantic comparison, experiment curation, and deployment-ready qualitative review.

Fractional vegetation cover project
RGB Multispectral CNN segmentation

Fractional Vegetation Cover Mapping

CNN-based semantic segmentation of UAS RGB and multispectral imagery for mapping fractional vegetation cover in semi-arid ecosystems.

Hyperspectral Semantic segmentation Workflow

FVC Function Viewer

End-to-end workflow for fractional vegetation cover semantic segmentation from UAS imagery, including preprocessing, label harmonisation, UNet-based modelling, and spectral attention experiments across multispectral and hyperspectral inputs.

Full VNIR-SWIR range: 420 – 1600 nm