Glacial lakes provide insight into the melting rates of glaciers; thus, the ability to automatically detect and map them opens possibilities for improved monitoring of the changing size of glacial lakes. An accurate automated method for glacial lake segmentation would provide the means to perform constant monitoring without the need for tedious manual labeling. This work utilizes a deep learning approach using semantic segmentation in MATLAB with convolutional neural networks (CNNs) to automatically detect and map glacial lakes. This work can be used to produce quick estimates of lake area in order to monitor changes in their size. The CNN used is DeepLab v3+ with a Resnet 18 backbone. The algorithm correctly identified 93.322 % of lake pixels and has a mean BF score of 0.98652, meaning that the generated boundaries closely match the truth. These results show that this is a viable method for the detection and mapping of glacial lakes.
Deep Learning and Semantic Segmentation for the Detection and Mapping of Glacial Lakes
2024-07-15
2616433 byte
Conference paper
Electronic Resource
English
Deep Learning on 3D Point Cloud for Semantic Segmentation
Springer Verlag | 2021
|Deep Learning on 3D Point Cloud for Semantic Segmentation
TIBKAT | 2022
|