Remote Sensing Research Provides New Insights into Land Damage Monitoring

Researchers at Xinjiang University have made significant strides in land damage monitoring using remote sensing techniques. The study, focusing on the middle section of the Tianshan Mountains in Xinjiang, China, aimed to develop an automatic classification model of remote sensing image land damage. This model, based on the SENetV2-COT-DeepLabV3, integrated contextual transformer and SENetV2 modules to enhance feature extraction and segmentation abilities. The research demonstrated improved model performance, with recognition accuracy exceeding 85%, and identified land damage types in mining areas with high precision and efficiency.

Key Takeaways:

• The research developed an automatic classification model of remote sensing image land damage, SENetV2-COT-DeepLabV3, to enhance feature extraction and segmentation abilities.
• The model was trained on a dataset of 59198 samples, extended to 177594 samples through data augmentation, and demonstrated recognition accuracy exceeding 85%.
• Comparative experiments showed that the improved model outperformed mainstream models, such as FCN and PSPNet, in four indicators: MIoU, mRecall, mPrecision, and mDice.
• The study proposed a deep learning remote sensing interpretation system for land damage types in mining areas, providing an intelligent solution for dynamic monitoring and ecological restoration management.
• The research has significant implications for the coordinated development of mine development and environmental protection.

Statistics:

• 59198 samples were initially constructed for the middle section of the Tianshan Mountains.
• Data augmentation extended the sample set to 177594 samples.
• The model's segmentation accuracy was 1.63%-2.34% higher than that of DeepLabV3.
• Recognition accuracy exceeded 85%.
• Comparative experiments showed improved performance in four indicators: MIoU, mRecall, mPrecision, and mDice.

Sources:

• NewsRx research report, August 4, 2025 (NewsRx LLC).
• "Build and application of automatic classification model of land damage in the middle section of Tianshan Mountains." Shuiwen dizhi gongcheng dizhi, 2025,52(4):26-38.
• Editorial Office of Hydrogeology & Engineering Geology.