This is a project which is currently making use of HPC facilities at Newcastle University. It is active.
For further information about this project, please contact:
This project uses HPC to process large volumes of traffic camera imagery from across England in order to monitor vegetation phenology (seasonal changes such as leaf-out and senescence). The work combines image processing, machine learning and geospatial analysis to extract vegetation indices and phenological metrics from road-side cameras, and to compare these with satellite-based products.
On the HPC system, we will run image filtering, clustering and object and instance segmentation to identify vegetation. The resulting time series will be used to characterise spatial and temporal patterns of vegetation dynamics at regional to national scales.
The project primarily uses Python and R. Core Python tools include PyTorch-based YOLO models for object and instance segmentation, OpenCV and scikit-image for image preprocessing and feature extraction, NumPy/Pandas for data handling, and HDF5/NetCDF for intermediate storage. R is used for time-series analysis, statistical modelling and visualisation (tidyverse, ggplot2, etc.). Compute workloads are a mix of GPU-accelerated and CPU-intensive operations.