A multiscale feature fusion enhanced CNN with the multiscale channel attention mechanism for efficient landslide detection (MS2LandsNet) using medium-resolution remote sensing data

ABSTRACT

Deep learning (DL) models have been widely used for remote sensing-based landslide mapping due to their impressive capabilities for automatic information extraction. However, the large volumes of parameters and calculations have compromised the efficiency of DL models in extracting landslides from a large set of RS images. Lightweight convolutional neural networks (CNNs) exhibit promising feature representation abilities with fewer parameters. This study aims to introduce a new lightweight CNN called MS2LandsNet, designed to detect landslides with both high efficiency and accuracy. The MS2LandsNet consists of three down-sampling stages embedded with multi-scale feature fusion (MFF), aiming to decrease parameters while aggregating contextual features. Additionally, we incorporate multi-scale channel attention (MSCA) into MFF to improve performance. According to experimental results on three landslip datasets, MS2LandsNet obtains the highest F1 score of 85.90% and the highest IoU of 75.28%. Notably, MS2LandsNet accomplishes the resuts with the fewest parameters and the fastest inference speed, outperforming seven classical semantic segmentation models and three lightweight CNNs. The proposed lightweight model holds potential for application on a cloud computing platform for larger-scale landslide mapping tasks in future work.

KEYWORDS:

• Landslide detection
• multi-scale feature fusion
• attention mechanism
• convolutional neural network
• remote sensing

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Landslide4Sense dataset was acquired at the website https://www.iarai.ac.at/landslide4sense/ (assessed on August 1st 2023). Codes for data acquisition can be accessed via https://drive.google.com/file/d/1u7TF08iXF5RqPAjxZMGfn55IVGwGJ6Wr/view?usp=drive_link. Codes for data pre-processing can be accessed via https://drive.google.com/file/d/1SBAN6n1dQvY8Wf8AHenQ1PuZNCGWfj5Z/view?usp=sharing.

Additional information

Funding

This study was supported by the Network Security and Information Program of the Chinese Academy of Sciences under Grant CAS-WX2021SF-0106-02; the Second Tibetan Plateau Scientific Expedition and Research Program (STEP) under Grant 20190ZKK1006; the National Natural Science Foundation of China under Grant 42130508; and the Key Project of Innovation LREIS under Grant KPI011.