Thermal imaging and tree inspection: what research reports about detecting decay
A 2012 study of 48 felled trees in Singapore found trunk surface temperature followed bark damage and sun, not internal decay. A 2019 conference paper and a 2024 log study report cavities found under set conditions.
Checked against the sources at the bottom of this page on October 16, 2026. Rules, fees and pay change: the source has the last word.
Infrared cameras read the temperature of a trunk surface without touching the tree. The idea tested in the papers below is that decayed or hollow wood conducts and stores heat differently from sound wood, so a defect may show as a cooler or warmer patch on the bark. Three open papers test that idea, and their results differ. Other methods that look inside a stem are covered in the guide to sonic and acoustic tomography.
The Singapore study of 48 trees
Daniel Burcham, Eng-Choon Leong, Yok-King Fong and Puay-Yok Tan published their study in Arboriculture and Urban Forestry in November 2012. They photographed 48 Casuarina equisetifolia trees with an infrared camera on March 5, 2011, between 2 and 4 pm, from 10 m away. The trees were then felled and cut into sections to measure what was inside.
| Internal condition found | Share of the 48 trees |
|---|---|
| Decayed | 27 percent |
| Discolored | 62 percent |
| Cavitated by termite infestation | 6 percent |
| Undamaged | 2 percent |
The authors saw cooler spots on the bark over mechanical damage, detached bark, cracks, termite nests and shed branch wounds. They saw warmer patches on trees in direct sun at forest gaps and access roads. Trees in dense forest showed fairly even surface temperatures.
Internal condition explained a small part of the temperature variation, with r squared values of 0.001 to 0.096. Near internal defects the average deviation from the trunk's temperature trend was 0.2 degrees C. Near the marking tape on the trunks it was 1.5 degrees C. The authors conclude that no clear relationship was found between the extent of internal defects and surface temperature, and that the technique does not give accurate results about the internal condition of trees.
Earlier results the Singapore paper cites
The paper describes the first publications as reporting trunk surface temperatures that were variable and cooler over internal defects. It cites two later quantitative studies with narrower findings.
- Burcham and co-authors, 2012: internal defects covering at least 76 percent of the stem cross-sectional area were needed to produce a measurable drop in surface temperature under controlled conditions.
- Bellett-Travers and Morris, 2010: surface temperature correlated with radial wood thickness (r squared 0.31 to 0.55) only in trees with a distinct, compartmentalized cavity. The link was absent in healthy trees and in trees with incipient decay.
The Singapore authors read these as pointing to a narrow range of relatively large internal defects that the camera may find.
Hong Kong conference paper, 2019
A team at the Hong Kong Polytechnic University presented a method at the 40th Asian Conference on Remote Sensing in Daejeon in October 2019. It groups trunk areas with similar surface temperature by k-means clustering and then looks for large temperature changes with a Sobel gradient filter. The aim is to flag potential cavities without relying on visual reading of the image.
The method was tried on trees in four species groups, each with at least one tree known to have a structural defect and one healthy tree. The team also studied what changes the result: acquisition time, humidity, temperature, light intensity, weather, camera distance and bark roughness. They report that images taken in the optimized conditions identified the internal cavities, and that a certified arborist checked the target trees on site. The record is a conference presentation, not a journal article.
Sri Lanka log study, 2024
Researchers at the University of Peradeniya, writing in the proceedings of the RESCON 2024 congress, used tree logs as models. Logs differ in thermal profile from live trees, so the team cooled them with water at 10 to 15 degrees C and waited 10 to 20 minutes before taking a thermogram. They report that the method detected cavities and estimated their size and depth from the bark.
Larger cavities and shorter distances from the bark to the cavity raised the chance of detection. A longer camera distance improved image contrast by widening the field of view. The authors acknowledge that the analyses were case by case, given the volatility of the factors involved. The work covers logs only, not standing trees.
Where the three papers differ
The Singapore study tested standing trees in a coastal forest and found surface temperature tracking external features and sunlight. The Hong Kong paper and the Sri Lanka study report detection under chosen conditions: set capture settings in one, active cooling of logs in the other.
Sources
- Burcham, Leong, Fong and Tan, An Evaluation of Internal Defects and Their Effect on Trunk Surface Temperature in Casuarina equisetifolia, Arboriculture & Urban Forestry 38(6), November 2012, read October 16, 2026
- Kwok and others, Detection of structural tree defects using thermal infrared imaging, 40th Asian Conference on Remote Sensing (2019), PolyU Scholars Hub record, read October 16, 2026
- Baduge, Jayalath and Dasanayake, Qualitative study on detecting internal cavities of tree logs using infrared thermography, RESCON 2024, University of Peradeniya, read October 16, 2026