Purdue Team Introduces Advance in Automatic Forest Mapping Technology

WEST LAFAYETTE, Ind. — The way lightning travels from the sky to the ground inspired the concept behind a new algorithmic approach to digitally separating individual trees from their forests in automatic forest mapping.

“As lightning travels from the sky to the ground, it finds the path of least resistance through the atmosphere,” said Joshua CharpentierPhD student at Purdue’s Lyles School of Civil Engineering. This led him to think similarly about his digital forest data, or point cloud.

“If I could somehow treat all the points in this scatter plot as a path of least resistance, that would tell me something about where the tree is,” Carpenter said. The concept also works from a plant biology perspective.

“Each leaf of a tree needs to be supplied with nutrients, and nutrients come from the soil. Thus, we find the shortest path for tree nutrients from the canopy to the ground.

Carpenter and four Purdue co-authors published the details of their mapping methods recently in the journal Remote Sensing. The approach means the difference between mapping a few trees and mapping hundreds of acres both quickly and with great precision. It could also lead to the creation of digital twins of forests, which could improve management planning in the face of climate change, epidemics and population growth.

This image shows the input and output data of the tree segmentation algorithm. The input data (left) is colored by elevation. The algorithm results (right) use color to segment each tree from the point cloud. (Photo Purdue University/Joshua Carpenter) Download image

The work was partially supported by Purdue’s Integrated Digital Forestry Initiative. This initiative, one of Purdue’s Next Moves five strategic investments, leverages digital technology and multidisciplinary expertise to measure, monitor and manage urban and rural forests to maximize social, economic and ecological benefits.

“We have developed a new individual tree segmentation algorithm that can be used to inventory trees over large areas,” the paper’s co-author said. Jinha Jungassistant professor of public works. Carpenter is a member of Jung’s Geospatial Data Science Laboratoryspecialized in mapping and measurement.

“Another contribution of this paper is how to evaluate the performance of the segmentation algorithm with data collected in the field,” Jung said.

The algorithm was found to be more accurate by most measures, often by far, compared to the current state of the art. Validation involves directly marking and measuring individual trees in the field to correlate with LiDAR data collected at ground level and from the air at different times of the year to capture leafy and leafless trees.

The team is still tackling issues with its three data collection methods: photogrammetry (creating 3D images from 2D photographs) and two types of LiDAR (aerial and ground-based).

The scatter plot data has the same structure, but the data from each method contains different anomalies. One can capture the details of the top of the tree canopy quite well but miss elements of the trunk and vice versa. Sometimes landscape features also block data collection.

“The goal is to use all the different point clouds available to create a flexible algorithm,” Carpenter explained. “But finding a method to work with each of the specific abnormalities is a challenge.”

Working in the 400-acre Martell Forest about 8 miles east of campus, the Purdue team continues to expand the reach of its technology.

“How can we go from several hundred hectares to several thousand or several hundred thousand, then to all the trees on the planet? This is the future,” said the co-author of the article Songlin FeiProfessor and Dean’s Chair in Remote Sensing at Forestry and natural resources. “The problem is how to scale it.”

Inventory requires tedious fieldwork to sample 5% or 10% of an area. “100% inventory was never an option. This article presents technologies that allow a census of each tree. We are talking about a huge leap forward,” Fei said.

The remote sensing paper focuses on forest mapping, but more algorithms will be needed to perform comprehensive inventories.

“We can make diameter measurements with this data. But what about other key characteristics of the inventory, such as straightness, wood quality or species identification? These have yet to be accomplished,” Fei said.

Technologies now make it possible to produce a digital twin of an entire forest to see the potential effects of an ice storm or high winds.

“If you do a forest management plan, you can’t just harvest the trees and see what it looks like,” Fei noted. “But in the digital world, you can cut down any tree you want, and you can put it back. This allows you to do simulations and better plan management.

Over the past few decades, geospatial data has dramatically increased agricultural production. Purdue researchers seek to do the same for forestry, an important source of raw materials for construction and fuel. Catastrophic wildfires and invasive species that have wiped out vast stands of chestnut and American ash are now drawing attention to the importance of forests.

“We’ve applied all of these technologies successfully to agriculture,” Carpenter said. “But other areas now need our attention.”

Writer: Steve Kopes

Media contact: Maureen Manier, [email protected]
Sources: Joshua Carpenter, [email protected]; Jinha Jung, [email protected]; Songlin Fei, [email protected]

Agricultural communications: 765-494-8415;

Maureen Manier, Head of Department, [email protected]

Agricultural news page

Comments are closed.