Extraction of Forest Road Information from CubeSat Imagery Using Convolutional Neural Networks ...

• Main Authors: Winiwarter, Lukas, Coops, Nicholas C., Bastyr, Alex, Roussel, Jean-Romain, Zhao, Daisy Q. R., Lamb, Clayton T., Ford, Adam T.
• Format: Text
• Language: English
• Published: Multidisciplinary Digital Publishing Institute 2024
• Subjects: Alces alces; Rangifer tarandus
• Online Access: https://dx.doi.org/10.14288/1.0441475; https://doi.library.ubc.ca/10.14288/1.0441475

Forest roads provide access to remote wooded areas, serving as key transportation routes and contributing to human impact on the local environment. However, large animals, such as bears (Ursus sp.), moose (Alces alces), and caribou (Rangifer tarandus caribou), are affected by their presence. Many publicly available road layers are outdated or inaccurate, making the assessment of landscape objectives difficult. To address these gaps in road location data, we employ CubeSat Imagery from the Planet constellation to predict the occurrence of road probabilities using a SegNet Convolutional Neural Network. Our research examines the potential of a pre-trained neural network (VGG-16 trained on ImageNet) transferred to the remote sensing domain. The classification is refined through post-processing, which considers spatial misalignment and road width variability. On a withheld test subset, we achieve an overall accuracy of 99.1%, a precision of 76.1%, and a recall of 91.2% (F1-Score: 83.0%) after considering these ...