Assessment of Tree Detection Methods in Multispectral Aerial Images

Dagoberto Pulido,Sertac Karaman,Joaquín Salas,Klaus Puettmann,Matthias Rös

doi:10.3390/rs12152379

Abstract

Detecting individual trees and quantifying their biomass is crucial for carbon accounting procedures at the stand, landscape, and national levels. A significant challenge for many organizations is the amount of effort necessary to document carbon storage levels, especially in terms of human labor. To advance towards the goal of efficiently assessing the carbon content of forest, we evaluate methods to detect trees from high-resolution images taken from unoccupied aerial systems (UAS). In the process, we introduce the Digital Elevated Vegetation Model (DEVM), a representation that combines multispectral images, digital surface models, and digital terrain models. We show that the DEVM facilitates the development of refined synthetic data to detect individual trees using deep learning-based approaches. We carried out experiments in two tree fields located in different countries. Simultaneously, we perform comparisons among an array of classical and deep learning-based methods highlighting the precision and reliability of the DEVM.

Highlights

Programs to reduce emissions from deforestation and forest degradation (e.g., REDD+ [1]) intend to mitigate the effects of climate change by providing forest landowners with economic incentives reflecting the value of the carbon stored within the trees
Our results show that Convolutional Neural Network (CNN)-based methods have become the leading performer
We describe how these images were processed to generate the input for the convolutional neural networks we utilize in this study

Summary

Introduction

Programs to reduce emissions from deforestation and forest degradation (e.g., REDD+ [1]) intend to mitigate the effects of climate change by providing forest landowners with economic incentives reflecting the value of the carbon stored within the trees. Despite advancements in remote sensing technology, manual labor still needs to accomplish many measurements, such as estimating the overall vegetation biomass and the carbon stored in individual trees and forests. It is common for field crews to travel to inventory plots and perform tasks such as counting and measuring tree sizes using visual estimations and manual measurements. This approach requires a considerable amount of time and resources, e.g., the USDA Forest Service spends more than 75% of the inventory costs on data collection [2]. We utilized the DSM, DTM, and NDVI to obtain a Digital Elevated Vegetation Model (DEVM). Classic approaches remain competitive and may offer advantages in settings where data collection and available computing resources for training are an issue

Methods

Results

Discussion

Conclusion