Mingzhi Zhang
Mingzhi Zhang has a total of 53 co-authors. Here are the top co-authors by number of shared publications:
- Hongyong Yuan
School of Safety Science, Tsinghua University, Beijing 100084, China
2 shared publications - Pingping Huang
College of Information Engineering, Inner Mongolia University of Technology, Hohhot 010080, China
2 shared publications - Weixian Tan
College of Information Engineering, Inner Mongolia University of Technology, Hohhot 010080, China
2 shared publications Yaolong Qi
Electronics Information Engineering. Beihang University,Beijing,China
2 shared publications- Baohe Wu
Institute of Exploration Technology, Chinese Academy of Geological Sciences, Chengdu 611734, China
1 shared publication - Chun Bao
School of Optics and Photonics, Beijing Institute of Technology, Beijing 100081, China
1 shared publication
Latest Publications
- Open Access
- 11 May 2026
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 15 (5) , 209
Cross-Modal Dynamic Feature Fusion for Visible-Infrared Debris Flow Segmentation
Gully type debris flows are sudden, highly destructive geological hazards requiring accurate, real-time monitoring for effective early warning. However, single-modal visual monitoring is sensitive to complex environments, while existing multi-modal fusion approaches rely on static strategies, limiting adaptability and modal complementarity. Blurred boundary segmentation, class imbalance, and real-time deployment challenges also remain unaddressed. To overcome these issues, this study proposes a cross-modal dynamic feature fusion framework integrating visible and infrared imagery, consisting of a shared backbone for multi-scale feature extraction, a dynamic feature aggregation module for adaptive heterogeneous fusion, a lightweight context-aware semantic segmentation network, and a composite loss function to enhance boundary delineation and mitigate class imbalance. Validated on a self-constructed dual-modal debris flow dataset and public benchmarks, the method achieves an mIoU of 75.6%, outperforming state-of-the-art methods by 3.1%. It meets real-time monitoring requirements and exhibits strong generalization, providing a practical solution for debris flow monitoring with great potential for disaster early warning deployment.
- 11 May 2026
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 15 (5) , 209
Open AccessCross-Modal Dynamic Feature Fusion for Visible-Infrared Debris Flow Segmentation
Gully type debris flows are sudden, highly destructive geological hazards requiring accurate, real-time monitoring for effective early warning. However, single-modal visual monitoring is sensitive to complex environments, while existing multi-modal fusion approaches rely on static strategies, limiting adaptability and modal complementarity. Blurred boundary segmentation, class imbalance, and real-time deployment challenges also remain unaddressed. To overcome these issues, this study proposes a cross-modal dynamic feature fusion framework integrating visible and infrared imagery, consisting of a shared backbone for multi-scale feature extraction, a dynamic feature aggregation module for adaptive heterogeneous fusion, a lightweight context-aware semantic segmentation network, and a composite loss function to enhance boundary delineation and mitigate class imbalance. Validated on a self-constructed dual-modal debris flow dataset and public benchmarks, the method achieves an mIoU of 75.6%, outperforming state-of-the-art methods by 3.1%. It meets real-time monitoring requirements and exhibits strong generalization, providing a practical solution for debris flow monitoring with great potential for disaster early warning deployment.
- Open Access
- 29 November 2024
- Published by MDPI AG in Sustainability
- Vol. 16 (23) , 10505
Cost Estimation for the Operation and Maintenance of Automated Monitoring and Early-Warning Equipment for Geological Hazards
Geological hazards impede regional economy sustainability. To limit their destructive impacts on human life and property, the Chinese government has independently developed automated monitoring and early-warning equipment, which has been deployed in over 250,000 locations nationwide, yielding effective early warnings. The smooth operation of this equipment necessitates substantial human, material, and financial resources for its maintenance. To allocate funds rationally, the Ministry of Finance of China has mandated the urgent establishment of budget standards for the operation and maintenance of automated monitoring and early-warning systems for geological hazards. Addressing the research gap in this area, this study meticulously develops a cost model, subcategorizing operating costs, maintenance costs, and management costs. Addressing the intricate issue of maintenance expenditures, this study ingeniously breaks down routine operations and urgent repairs stipulated in technical standards into personnel, materials, and vehicular needs for each equipment type. Considering the total manpower involved in equipment maintenance, the per-unit maintenance cost is determined. This method allocates costs to individual pieces of equipment, thereby sidestepping the quantification hurdle created by varying types and quantities of monitoring equipment at each monitoring site due to various geological disaster types and magnitudes, and technical personnel’s maintenance responsibility for multiple equipment types in a single operation. Finally, incorporating regional adjustment coefficients, we have formulated theoretical costs for the operation and maintenance of automated monitoring and early-warning equipment for geological hazards. By contrasting theoretical costs with actual project budgets, the error margin is within 2%. Following nationwide consultation, these theoretical costs have been officially endorsed as the budget standard. These standards will lay the groundwork for project budgeting and review, facilitate efficient fund utilization, and ensure the financial sustainability of monitoring and warning systems for geological hazards. Concurrently, this paper bridges the global lack in budget norms for the operation and upkeep of automated geological disaster monitoring systems. The cost calculation model introduced serves as a pivotal reference globally for the evaluation of analogous system’s operations and maintenance expenses.
- 29 November 2024
- Published by MDPI AG in Sustainability
- Vol. 16 (23) , 10505
Open AccessCost Estimation for the Operation and Maintenance of Automated Monitoring and Early-Warning Equipment for Geological Hazards
Geological hazards impede regional economy sustainability. To limit their destructive impacts on human life and property, the Chinese government has independently developed automated monitoring and early-warning equipment, which has been deployed in over 250,000 locations nationwide, yielding effective early warnings. The smooth operation of this equipment necessitates substantial human, material, and financial resources for its maintenance. To allocate funds rationally, the Ministry of Finance of China has mandated the urgent establishment of budget standards for the operation and maintenance of automated monitoring and early-warning systems for geological hazards. Addressing the research gap in this area, this study meticulously develops a cost model, subcategorizing operating costs, maintenance costs, and management costs. Addressing the intricate issue of maintenance expenditures, this study ingeniously breaks down routine operations and urgent repairs stipulated in technical standards into personnel, materials, and vehicular needs for each equipment type. Considering the total manpower involved in equipment maintenance, the per-unit maintenance cost is determined. This method allocates costs to individual pieces of equipment, thereby sidestepping the quantification hurdle created by varying types and quantities of monitoring equipment at each monitoring site due to various geological disaster types and magnitudes, and technical personnel’s maintenance responsibility for multiple equipment types in a single operation. Finally, incorporating regional adjustment coefficients, we have formulated theoretical costs for the operation and maintenance of automated monitoring and early-warning equipment for geological hazards. By contrasting theoretical costs with actual project budgets, the error margin is within 2%. Following nationwide consultation, these theoretical costs have been officially endorsed as the budget standard. These standards will lay the groundwork for project budgeting and review, facilitate efficient fund utilization, and ensure the financial sustainability of monitoring and warning systems for geological hazards. Concurrently, this paper bridges the global lack in budget norms for the operation and upkeep of automated geological disaster monitoring systems. The cost calculation model introduced serves as a pivotal reference globally for the evaluation of analogous system’s operations and maintenance expenses.
- Open Access
- 1 January 2024
- Published by China Geology in China Geology
- Vol. 7 (2) , 351-364
Short-term displacement prediction for newly established monitoring slopes based on transfer learning
- 1 January 2024
- Published by China Geology in China Geology
- Vol. 7 (2) , 351-364
Open AccessShort-term displacement prediction for newly established monitoring slopes based on transfer learning
- 15 September 2023
- Published by Springer Nature in Landslides
- Vol. 21 (1) , 71-81
Experimental investigation and field application of acoustic emission array for landslide monitoring
- 15 September 2023
- Published by Springer Nature in Landslides
- Vol. 21 (1) , 71-81
Experimental investigation and field application of acoustic emission array for landslide monitoring
- 30 August 2023
- Published by Springer Nature in Landslides
- Vol. 20 (12) , 2709-2723
InSAR-based landslide detection method with the assistance of C-index
- 30 August 2023
- Published by Springer Nature in Landslides
- Vol. 20 (12) , 2709-2723
InSAR-based landslide detection method with the assistance of C-index
- Open Access
- 1 October 2020
- Published by IOP Publishing in IOP Conference Series: Earth and Environmental Science
- Vol. 570 (6) , 062048
Research, development, and field trial of the universal Global Navigation Satellite System receivers
Global Navigation Satellite System (GNSS) can provide continuous, high precision, and absolute displacement information on landslides. It can thus be used in landslide monitoring applications. However, the cost of high-precision GNSS is too high for large-scale deployment. From an application perspective, we reviewed the status and challenges of applying GNSS technology to landslide monitoring. We then investigated the data specification for a landslide monitoring GNSS receiver to better meet specified engineering requirements. We further developed prototype universal GNSS monitoring devices and deployed them in the Heitai area for testing. Approximately two months of testing indicated that the prototype universal GNSS devices well meet the engineering requirements, and the tested GNSS data are in good agreement with the crack data. The hourly monitoring GNSS solution achieves sub-millimeter level accuracy and can be used for extracting landslide information.
- 1 October 2020
- Published by IOP Publishing in IOP Conference Series: Earth and Environmental Science
- Vol. 570 (6) , 062048
Open AccessResearch, development, and field trial of the universal Global Navigation Satellite System receivers
Global Navigation Satellite System (GNSS) can provide continuous, high precision, and absolute displacement information on landslides. It can thus be used in landslide monitoring applications. However, the cost of high-precision GNSS is too high for large-scale deployment. From an application perspective, we reviewed the status and challenges of applying GNSS technology to landslide monitoring. We then investigated the data specification for a landslide monitoring GNSS receiver to better meet specified engineering requirements. We further developed prototype universal GNSS monitoring devices and deployed them in the Heitai area for testing. Approximately two months of testing indicated that the prototype universal GNSS devices well meet the engineering requirements, and the tested GNSS data are in good agreement with the crack data. The hourly monitoring GNSS solution achieves sub-millimeter level accuracy and can be used for extracting landslide information.
- Open Access
- 29 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 13 (19391404) , 6249-6259
Multidimensional Response Evaluation of Remote-Sensing Vegetation Change to Drought Stress in the Three-River Headwaters, China
The impact of drought on ecosystems has become increasingly prominent. This study used the MODIS remote-sensing vegetation index and standardized precipitation index to quantify vegetation coverage status and meteorological drought level. Using the copula method, the joint distribution of the drought index and related vegetation cover variables was simulated for the first time. The conditional distribution of vegetation biomass was deduced, and the multidimensional response between vegetation biomass and drought was explored to understand the possible vegetation loss under different drought severity conditions. The Three-River Headwaters (TRH) region is the source place of the Yangtze River, Yellow River, and Lancang River, where the ecosystem is characterized by innate vulnerability. Using this region as a case study, the results show that the spatiotemporal evolution of drought and vegetation in the TRH region has apparent regional heterogeneity. The vegetation cover in the eastern region is significantly better than that in the northwestern region, while the vegetation growth trend in the northwestern region is stronger than that in the southeastern region. It is feasible to build a vegetation-drought multidimensional response model based on the copula method. In 92.5% of the TRH region, vegetation cover was significantly affected by the severity of the drought. The impact on the growth of vegetation caused by persistent drought events was higher than that of short-term but high-intensity drought. The vegetation in the TRH area has a certain degree of drought resistance. This study provides outstanding theoretical support and reference for the protection of the TRH basin ecosystem.
- 29 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 13 (19391404) , 6249-6259
Open AccessMultidimensional Response Evaluation of Remote-Sensing Vegetation Change to Drought Stress in the Three-River Headwaters, China
The impact of drought on ecosystems has become increasingly prominent. This study used the MODIS remote-sensing vegetation index and standardized precipitation index to quantify vegetation coverage status and meteorological drought level. Using the copula method, the joint distribution of the drought index and related vegetation cover variables was simulated for the first time. The conditional distribution of vegetation biomass was deduced, and the multidimensional response between vegetation biomass and drought was explored to understand the possible vegetation loss under different drought severity conditions. The Three-River Headwaters (TRH) region is the source place of the Yangtze River, Yellow River, and Lancang River, where the ecosystem is characterized by innate vulnerability. Using this region as a case study, the results show that the spatiotemporal evolution of drought and vegetation in the TRH region has apparent regional heterogeneity. The vegetation cover in the eastern region is significantly better than that in the northwestern region, while the vegetation growth trend in the northwestern region is stronger than that in the southeastern region. It is feasible to build a vegetation-drought multidimensional response model based on the copula method. In 92.5% of the TRH region, vegetation cover was significantly affected by the severity of the drought. The impact on the growth of vegetation caused by persistent drought events was higher than that of short-term but high-intensity drought. The vegetation in the TRH area has a certain degree of drought resistance. This study provides outstanding theoretical support and reference for the protection of the TRH basin ecosystem.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 3, 1-6
Slope Sliding Direction Estimate Based on Ground-based D-InSAR
Landslide is a common natural geological disaster, and the development process of landslide is so complicated, then makes the prevention and control of landslide very difficult. In the process of landslide monitoring and control, the estimation of slope sliding direction is very important. Based on the monitoring data of D-InSAR (Differential Intereferometric Synthetic Aperture Radar), combined with the geometric relationship between the monitoring target and radar, as well as the objective laws observed in the natural deformation process of the slope, a method to estimate the sliding direction of the slope is proposed in this paper. This paper uses the monitoring data of two slopes as an example and verifies the feasibility and effectiveness of the method in the face of the point target and the whole slope through comparison with the monitoring data of GPS, the comparison results show good consistency.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 3, 1-6
Slope Sliding Direction Estimate Based on Ground-based D-InSAR
Landslide is a common natural geological disaster, and the development process of landslide is so complicated, then makes the prevention and control of landslide very difficult. In the process of landslide monitoring and control, the estimation of slope sliding direction is very important. Based on the monitoring data of D-InSAR (Differential Intereferometric Synthetic Aperture Radar), combined with the geometric relationship between the monitoring target and radar, as well as the objective laws observed in the natural deformation process of the slope, a method to estimate the sliding direction of the slope is proposed in this paper. This paper uses the monitoring data of two slopes as an example and verifies the feasibility and effectiveness of the method in the face of the point target and the whole slope through comparison with the monitoring data of GPS, the comparison results show good consistency.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Rotating micro deformation monitoring radar
Ground-based radar is an application of microwave imaging on ground platform, which has the characteristics of repeatable observation, stability and controllability, and can be applied to the long-term observation of slope deformation, urban subsidence and landslides. However, the fundamental imaging principle of current ground-based radar is to create a large synthetic aperture through radar's moving on a linear orbit, but adopting this method, the angle of view is limited and 360° monitoring is impossible. Therefore, it is of great significance to study the full-azimuth micro deformation monitoring radar (namely the rotating micro deformation monitoring radar), for observing large scenes featuring with large angle of view.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Rotating micro deformation monitoring radar
Ground-based radar is an application of microwave imaging on ground platform, which has the characteristics of repeatable observation, stability and controllability, and can be applied to the long-term observation of slope deformation, urban subsidence and landslides. However, the fundamental imaging principle of current ground-based radar is to create a large synthetic aperture through radar's moving on a linear orbit, but adopting this method, the angle of view is limited and 360° monitoring is impossible. Therefore, it is of great significance to study the full-azimuth micro deformation monitoring radar (namely the rotating micro deformation monitoring radar), for observing large scenes featuring with large angle of view.