Analyzing Multispectral Satellite Imagery of South American Wildfires Using Deep Learning

Abstract

Since severe droughts are occurring more frequently and lengthening the dry season in the Amazon Rainforest, it is important to respond to active wildfires promptly and to forecast them before they become inextinguishable. Though computer vision researchers have applied algorithms on large databases to automatically detect wildfires, current models are computationally expensive and are not versatile enough for the low technology conditions of regions in South America. This comprehensive deep learning study first trains a Fully Convolutional Neural Network with skip connections on multispectral Landsat 8 images of Ecuador and the Galapagos. The model uses Green and Short-wave Infrared bands as inputs to predict each image's corresponding pixel-level binary fire mask. This model achieves a 0.962 validation F2 score and a 0.932 F2 score on test data from Guyana and Suriname. Afterward, image segmentation is conducted on the Cirrus Cloud band using K-Means Clustering to simplify continuous pixel values into three discrete classes representing the degree of cirrus cloud contamination. Two additional Convolutional Neural Networks are trained to classify the presence of a wildfire in a patch of land using these segmented cirrus images. The "experimental" model trained on the segmented inputs achieves 96.5% binary accuracy and has smoother learning curves than the "control model" that is not given the segmented inputs. This proof of concept reveals that feature simplification can improve the performance of wildfire detection models. Overall, the software built in this study is useful for early and accurate detection of wildfires in South America.

A Deep Learning Ensemble Framework for Off-Nadir Geocentric Pose Prediction

“…We trained a fully convolutional neural network to predict the elevation mask of each RGB image. The model was based on the U-Net architecture designed for a wildfire classification task, but included modifications such as having more convolutional blocks [5]. The filter size was 5x5 throughout the model.…”

“…However, these models underfitted the data and could not achieve an MAE less than 0.8 for both scale and angle. Thus, we concluded that the task at hand was sufficiently complex as to require a very deep network with greater than five million parameters; therefore, we focused our efforts towards this area, ultimately choosing a similar architecture that Sun, C. (2022) used for wildfire detection [5].…”

A Deep Learning Ensemble Framework for Off-Nadir Geocentric Pose Prediction

“…We trained a fully convolutional neural network to predict the elevation mask of each RGB image. The model was based on the U-Net architecture designed by Sun (2022) for a wildfire classification task, but included modifications such as having more convolutional blocks, and as a result, more skip connections between downsampling and upsampling layers [7]. The filter size was 5 × 5 throughout the model.…”