Researchers have developed an algorithm that uses computer vision techniques to accurately measure trees almost five times faster than traditional, manual methods. The University of Cambridge researchers developed the algorithm, which provides an accurate measurement of tree diameter, an important measurement used by scientists to monitor forest health and levels of carbon sequestration. Photo credit: University of Cambridge
Researchers have developed an algorithm that uses computer vision techniques to accurately measure trees almost five times faster than traditional, manual methods.
The University of Cambridge researchers developed the algorithm, which provides an accurate measurement of tree diameter, an important measurement used by scientists to monitor forest health and levels of carbon sequestration.
The algorithm uses low-cost, low-resolution LiDAR sensors built into many cellphones and delivers results that are just as accurate but much faster than manual measurement techniques. The results are published in the journal Remote Sensing.
The primary manual measurement used in forest ecology is tree diameter at breast height. These measurements are used to make statements about the health of trees and the broader forest ecosystem, and how much carbon is being sequestered.
Although this method is reliable, the method is time consuming as the measurements are taken tree by tree from the ground. In addition, human errors can lead to measurement deviations.
“When you’re trying to figure out how much carbon a forest is sequestering, these ground-based measurements are tremendously valuable, but also time-consuming,” said first author Amelia Holcomb of the Department of Computer Science and Technology in Cambridge. “We wanted to know if we could automate this process.”
Some aspects of forest measurement can be made with expensive specialty LiDAR sensors, but Holcomb and her colleagues wanted to see if those measurements could be made with cheaper, lower-resolution sensors like those used in some cellphone augmented reality applications.
Other researchers have done some forest measurement studies using this type of sensor, but these have focused on heavily managed forests where trees are tall and evenly spaced and undergrowth is regularly cleared. Holcomb and her colleagues wanted to test whether these sensors could provide accurate results for unmanaged forests quickly, automatically and in a single image.
“We wanted to develop an algorithm that could be used in more natural forests and that could deal with things like low-hanging branches or trees with natural irregularities,” Holcomb said.
Researchers designed an algorithm that uses a smartphone LiDAR sensor to automatically estimate trunk diameter from a single image under realistic field conditions. The algorithm has been integrated into a specially developed app for an Android smartphone and can provide results in near real time.
To develop the algorithm, the researchers first collected their own data set by manually measuring and photographing trees. Using image processing and computer vision techniques, they were able to train the algorithm to distinguish trunks from large branches, determine which way trees were leaning, and other information that could help it refine information about forests.
Researchers tested the app in spring, summer and autumn in three different forests – one each in the UK, US and Canada. The app was able to identify 100% of the logs and had an 8% mean error rate, which is comparable to the error rate when measuring by hand. However, the app significantly accelerated the process, about four and a half times faster than measuring trees manually.
“I was surprised the app worked so well,” said Holcomb. “Sometimes I like to challenge it with a particularly crowded patch of woods or a particularly odd-shaped tree, and I think there’s no way it’s going to get it right, but it does.”
Because their measuring tool requires no special training and uses sensors already built into an increasing number of phones, the researchers say it could be an accurate, low-cost tool for forest measurements, even in complex forest conditions.
The researchers plan to make their app publicly available for Android phones later this spring.
More information: Amelia Holcomb et al., Robust Single-Image Tree Diameter Estimation with Mobile Phones, Remote Sensing (2023). DOI: 10.3390/rs15030772
Provided by the University of Cambridge
Citation: Phone-Based Measurements Provide Fast, Accurate Information on Forest Health (2023 March 6) Retrieved March 6, 2023 from https://phys.org/news/2023-03-phone-based-fast-accurate -health-forests.html
This document is protected by copyright. Except for fair trade for the purpose of private study or research, no part may be reproduced without written permission. The content is for informational purposes only.
