Tree Diameter Estimation and Carbon Sequestration Quantification Using Deep Learning

Abstract

The quantification of Carbon Sequestration using deep learning project aims to simplify the estimation of a tree’s carbon uptake potential using machine learning. Since trees are vital carbon sinks, accurately estimating their sequestration potential plays a crucial role in combating climate change. Traditional methods are often manual and time-consuming, so this project leverages deep learning and image processing for faster and more scalable results. A Convolutional Neural Network (CNN) is employed to estimate the Diameter at Breast Height (DBH) from tree trunk images, a key parameter for calculating biomass and, in turn, carbon storage. The predicted DBH is then used in conjunction with standard forest equations to estimate carbon sequestration.
The model achieved a Mean Absolute Error (MAE) of 3.161cm, Root Mean Squared Error (RMSE) of 3.834 cm. and an overall accuracy of 89.87%, indicating strong predictive performance. The solution is deployed using a backend system and a user-friendly Streamlit frontend for image upload and real-time predictions. This tool benefits environmentalists, researchers, and policymakers by offering a fast, accurate, and accessible method for ecological monitoring and carbon offset planning. It promotes sustainability through the integration of modern technology.