收藏切换
Dynamic Diagnosis Method for Mine Reclamation Soil Quality Based on the Fusion of Swin-Unet and Grad-CAM
收藏切换
Junhuan Lu, Haojie Qin, Shuzhi Zhang, Yongjin Song
Mining Research and Development | 2026, 46(7) : 283 - 292
Less
收藏切换
Mining Research and Development | 2026, 46(7): 283-292
Dynamic Diagnosis Method for Mine Reclamation Soil Quality Based on the Fusion of Swin-Unet and Grad-CAM
Full
Junhuan Lu, Haojie Qin, Shuzhi Zhang, Yongjin Song
Affiliations
  • Guangxi Natural Resources Vocational and Technical College, Chongzuo, Guangxi 532199, China
Published: 2026-07-25
Outline
收藏切换

To address the problems that existing diagnosis methods can hardly capture the long-term variation law of mine soil quality, the evaluation of reclamation effect lags behind, and the soil quality diagnosis results lack interpretability, an improved dynamic diagnosis model for mine reclamation soil quality based on Swin-Unet and Gradient-Weighted Class Activation Mapping (Grad-CAM) visualization technology was proposed (IPPSUG model). Raw soil image features were extracted by the model via Swin-Unet, and noise was suppressed while key features were highlighted with the squeeze-and-excitation network module. Meanwhile, the model structure was optimized by the pruning algorithm to reduce computational complexity. Finally, the interpretability of diagnosis results was improved by Grad-CAM. The test results show that the accuracy and precision of soil quality diagnosis of the PSSU model reach 96.11% and 97.63%, respectively. The determination coefficient of IPSSUG model for identifying available potassium content is 0.983, and the recognition rate of soil-polluted areas is 98.40%. Test data are superior to those of the comparison models, which fully demonstrates the feasibility and superiority of this model. In addition, the degree of correlation between main influencing factors and soil quality indicators was quantified by IPSSUG model, the reliability of soil quality diagnosis results was improved, and a new method was provided for research on mine reclamation soil quality diagnosis.

Soil quality diagnosis  /  Squeeze-and-excitation network module  /  Pruning algorithm  /  Swin-Unet  /  Grad-CAM
卢俊寰, 覃豪杰, 张树志, 宋勇进. 基于Swin-Unet与Grad-CAM融合的矿山复垦土壤质量动态诊断方法. 矿业研究与开发, 2026 , 46 (7) : 283 -292 .
Junhuan Lu, Haojie Qin, Shuzhi Zhang, Yongjin Song. Dynamic Diagnosis Method for Mine Reclamation Soil Quality Based on the Fusion of Swin-Unet and Grad-CAM[J]. Mining Research and Development, 2026 , 46 (7) : 283 -292 .
Year 2026 volume 46 Issue 7
63
7
Cite this Article
BibTeX
Article Info
  • Receive Date:2025-08-22
  • Online Date:2026-10-27
  • Published:2026-07-25
Article Data
Affiliations
History
  • Received:2025-08-22
Funding
Affiliations
    Guangxi Natural Resources Vocational and Technical College, Chongzuo, Guangxi 532199, China
References
Share
https://castjournals.cast.org.cn/joweb/kyyjykf/EN/
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT