2021/03/20 by Kanchane Gunawardena · 1 voice
Environmental Science · #Urban Heat Island Mitigation #Remote Sensing in Agriculture #Remote Sensing and LiDAR Applications
paper · doi:10.5281/zenodo.19864210
openalex publication_date 2021/03/20 · openalex created_date 2026/04/29 · openalex updated_date 2026/07/14
This document describes the MATLAB workflow (updated to Python translation ThermogramExtractorv1.1), for extracting vegetation canopy temperatures from FLIR thermograms. Radiometric images yield direct temperature fields when available, while non-radiometric images are converted via on-image colourbar decoding with optional Optical Character Recognition (OCR) of colourbar endpoints. Given the tool’s development to assess building surface greening canopies, the AUTO mode allows for vegetation to be segmented from a companion colour photograph using a robust cascade of RGB-derived methods [ExG, VARI], HSV gating, and 2-cluster RGB k-means, with coverage-based acceptance and morphology refinement, while via the MANUAL mode a binary-mask-based extraction allows for the extraction of any data segment of interest. Parameters are input from an Excel workbook to enable reproducible, site-specific refinement (e.g., shadow compensation, threshold offsets). This is a supplementary method statement to a study by Gunawardena (2021, 2024) that implemented v1.0 of this tool (2021), which has been updated here as v1.1 (Python implementation), March 2026, TU Delft.