Semantic Segmentation of Mining-Induced Damage Using U-Net with Attention Mechanism
Abstract
thereby enabling scientific management of ecological restoration. This study utilizes GF-2 data and employs U-Net model, combined with
three attention mechanisms, for the high-level semantic extraction of land damage in three typical landforms. The results indicate that, for
imbalanced datasets from mining areas, the U-Net model integrated with attention mechanisms enable accurate semantic segmentation of land
damage types. The incorporation of scSE attention mechanism enhances the performance of U-Net model in distinguishing between different
types of land damage caused by mining activities.
Keywords
Full Text:
PDFReferences
[1] General Office of the Ministry of Natural Resources. Notice of the General Office of the Ministry of Natural Resources on carrying out
the pilot work for the survey of land damaged by mining across the country [Z]. Beijing, 2024-07-18.
[2] Cheng J, Deng C, Su Y, et al. Methods and Datasets on Semantic Segmentation for Unmanned Aerial Vehicle Remote Sensing Images:
A Review [J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2024, 211: 1-34.
[3] Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation [C]. Medical Image Computing
and Computer-Assisted Intervention(MICCAI 2015). Cham: Springer, 2015: 234-241.
[4] Sun F, Fang F, Wang R, et al. An Impartial Semi-Supervised Learning Strategy for Imbalanced Classification on VHR Images [J]. Sensors, 2020, 20 (22).
[5] He R, Wang R, Xu H, et al. Application of land damage information extraction in open-pit mines in Hubei province based on U-Net [J].
Resources Environment & Engineering, 2024, 38 (1): 100 110.
[6] Dong J. Focal Loss Improves the Model Performance on Multi-Label Image Classifications with Imbalanced Data [C]. Proceedings of
the 2nd International Conference on Industrial Control Network And System Engineering Research. Kuala Lumpur, Malaysia: Association for Computing Machinery, 2020: 18-21.
[7] Hu J, Shen L, Sun G, et al. Squeeze-and-Excitation Networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,
2017, 99.
[8] Roy A G, Navab N, Wachinger C. Concurrent Spatial and Channel Squeeze & Excitation in Fully Convolutional Networks [J]. Springer,
Cham, 2018.
[9] Woo S, Park J, Lee J Y, et al. CBAM: Convolutional Block Attention Module [J]. Springer, Cham, 2018.
[10] Bergstra J, Bengio Y. Random Search for Hyper-parameter Optimization [J]. The Journal of Machine Learning Research, 2012, 13: 281-
305.
DOI: http://dx.doi.org/10.70711/aitr.v3i5.8349
Refbacks
- There are currently no refbacks.
