An AI ‐based and coding‐free integration for forest Leaf Area Index calculation
Abstract Seasonal and spatial variations in Leaf Area Index (LAI, or plant area index, PAI) are challenging to delineate in tropical forests because the changes are often too subtle to be reliably detected. Many existing DHP software tools offer one‐click processing of all images using auto‐threshold segmentation (e.g. HemispheR, HemiPy and Hemisfer), but they produce results with large discrepancies. Some software (e.g. CAN‐EYE) requires manual tuning of each image, making large‐scale analysis impractical. We analysed 19,044 images from four tropical forest subtypes and found that using coding‐free AI software to process hemispherical images can significantly improve the consistency of canopy‐sky segmentation, thereby improving the detection of temporal and spatial variation of PAI. The results show that replacing the auto‐threshold segmentation with AI substantially reduced inter‐software disagreement, with the mean inter‐software range for PAI decreasing from 4.31 to 1.19, and delineating correct seasonal and spatial patterns. CAN‐EYE could identify the correct seasonal patterns but produced less accurate results than the CAN‐EYE‐AI integrated approach due to subjective bias. The high consistency achieved after AI integration enables reliable cross‐site and temporal comparisons. As users can combine the AI segmentation with other DHP software, our integrated, affordable and coding‐free method offers wide applicability and high consistency of PAI measurements, facilitating the advancement of tropical forest monitoring and research.
Authors
- Shalom Daniel Addo-Danso (ORCID: https://orcid.org/0000-0001-5074-1673)
- Mark Arcebal K. Naive (ORCID: https://orcid.org/0000-0002-1548-9465)
- H. Zhang (ORCID: https://orcid.org/0000-0003-4801-8771)
- Yadvinder Singh Malhi (ORCID: https://orcid.org/0000-0002-3503-4783)
- Minxue Tang (ORCID: https://orcid.org/0000-0003-1746-8411)
- Imma Oliveras Menor (ORCID: https://orcid.org/0000-0001-5345-2236)
- Stephen Adu‐Bredu (ORCID: https://orcid.org/0000-0001-7487-7911)
- Scott J. Davidson (ORCID: https://orcid.org/0000-0001-8327-2121)
- Akwasi Duah‐Gyamfi (ORCID: https://orcid.org/0000-0003-3030-063X)
- Sam Moore (ORCID: https://orcid.org/0000-0001-9504-5989)
- Cecilia A. L. Dahlsjö (ORCID: https://orcid.org/0000-0003-3795-1523)
- Emmanuel Amponsah Manu
- Forzia Ibrahim (ORCID: https://orcid.org/0000-0002-1156-2311)
- Zhiyuan Zhang (ORCID: https://orcid.org/0000-0002-2340-7001)
- Sophie Fauset (ORCID: https://orcid.org/0000-0003-4246-1828)
- William Hagan Brown
- Tao Ma
- Jill Ashley
- Fangni Liu
Institutions
- Centre National de la Recherche Scientifique (FR)
- Centre de Coopération Internationale en Recherche Agronomique pour le Développement (FR)
- Université du Québec à Montréal (CA)
- Chinese Academy of Sciences (CN)
- Université de Montpellier (FR)
- Xishuangbanna Tropical Botanical Garden (CN)
- University of Oxford (GB)
- Institut National de Recherche pour l'Agriculture, l'Alimentation et l'Environnement (FR)
- UMR Botanique et Modélisation de l’Architecture des Plantes et des végétations (FR)
- Council for Scientific and Industrial Research (GH)
- Institut de Recherche pour le Développement (FR)
- Forestry Research Institute of Ghana (GH)
- Imperial College London (GB)
- China Agricultural University (CN)
- Fujian Agriculture and Forestry University (CN)
- University of Birmingham (GB)
- University of Plymouth (GB)
- Umeå University (SE)
Publication Details
- Journal
- Methods in Ecology and Evolution
- Published
- 2026-10-26
- DOI
- https://doi.org/10.1111/2041-210x.70438
- Primary Topic
- Remote Sensing and LiDAR Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Natural Environment Research Council