MCD-Net: A Lightweight Deep Learning Baseline for Optical-Only Moraine Segmentation

TL;DR

This paper introduces MCD-Net, a lightweight deep learning model for moraine segmentation using optical imagery, supported by a large-scale annotated dataset, achieving high accuracy with reduced computational cost.

Contribution

The study presents the first large-scale optical-only moraine segmentation dataset and a lightweight, effective deep learning baseline model for glacial feature mapping.

Findings

MCD-Net achieves 62.3% mIoU and 72.8% Dice coefficient.

Reduces computational cost by over 60% compared to deeper models.

Optical imagery alone can reliably segment moraines despite some limitations.

Abstract

Glacial segmentation is essential for reconstructing past glacier dynamics and evaluating climate-driven landscape change. However, weak optical contrast and the limited availability of high-resolution DEMs hinder automated mapping. This study introduces the first large-scale optical-only moraine segmentation dataset, comprising 3,340 manually annotated high-resolution images from Google Earth covering glaciated regions of Sichuan and Yunnan, China. We develop MCD-Net, a lightweight baseline that integrates a MobileNetV2 encoder, a Convolutional Block Attention Module (CBAM), and a DeepLabV3+ decoder. Benchmarking against deeper backbones (ResNet152, Xception) shows that MCD-Net achieves 62.3% mean Intersection over Union (mIoU) and 72.8% Dice coefficient while reducing computational cost by more than 60%. Although ridge delineation remains constrained by sub-pixel width and spectral ambiguity, the results demonstrate that optical imagery alone can provide reliable moraine-body segmentation. The dataset and code are publicly available at https://github.com/Lyra-alpha/MCD-Net, establishing a reproducible benchmark for moraine-specific segmentation and offering a deployable baseline for high-altitude glacial monitoring.