Henggang Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Henggang Zhang has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
- Chenhao Chai
Key Laboratory of Mine Spatio-Temporal Information and Ecological Restoration, Henan Polytechnic University, Jiaozuo 454000, China
2 shared publications - Chenhui Zhu
College of Geography and Remote Sensing Science, Xinjiang University, Urumqi 830046, China
2 shared publications - Fan Feng
School of Surveying and Mapping, Information Engineering University, Zhengzhou 450001, China
2 shared publications - Hongtao Mu
School of Surveying and Mapping, Information Engineering University, Zhengzhou 450001, China
2 shared publications - Kaiyue Luo
College of Surveying and Geoinformatics, Tongji University, Shanghai 200070, China
2 shared publications - Tianyu Jiao
College of Geography and Remote Sensing Science, Xinjiang University, Urumqi 830046, China
2 shared publications
Latest Publications
- Open Access
- 19 September 2026
- Published by MDPI AG in Remote Sensing
- Vol. 18 (18) , 3220
Boundary-Aware Attention-Enhanced DeepLabV3+ for Landslide Segmentation Along the Sichuan–Tibet Transportation Corridor
Accurate delineation of landslides from high-resolution optical imagery is important for hazard assessment and infrastructure safety in mountainous regions. However, automatic segmentation remains challenging because landslides vary greatly in scale, morphology, surface appearance, and boundary clarity. These difficulties are particularly evident for old landslides, whose surface characteristics may become less distinct after long-term erosion, vegetation recovery, and surface modification. To establish a consistent baseline, five representative semantic segmentation models were evaluated on the Landslide Recognition along the Sichuan–Tibet Transportation Corridor (LRSTTC) dataset under a unified experimental protocol. DeepLabV3+ showed the strongest overall baseline performance and was selected as the base architecture. We then developed a Boundary-Aware Attention-Enhanced DeepLabV3+ (BAA-DeepLabV3+) framework, in which a Convolutional Block Attention Module (CBAM) is applied after Atrous Spatial Pyramid Pooling (ASPP) for high-level feature refinement, while an auxiliary boundary branch provides additional spatial supervision during training. In the controlled ablation experiment with seed 42, BAA-DeepLabV3+ improved the Intersection over Union (IoU) from 0.3644 to 0.4229 and the Dice coefficient from 0.5341 to 0.5944, with only a negligible increase in model parameters. Across three random seeds, the mean IoU increased from 0.3599 ± 0.0275 to 0.3876 ± 0.0315. External validation on the independent Bijie landslide dataset further improved the mean IoU from 0.7295 to 0.7444, with BAA-DeepLabV3+ outperforming the baseline under all three random seeds. These results indicate that attention-based feature refinement and boundary-aware supervision can work complementarily to improve landslide segmentation while maintaining a favorable balance between segmentation performance and computational efficiency.
- 19 September 2026
- Published by MDPI AG in Remote Sensing
- Vol. 18 (18) , 3220
Open AccessBoundary-Aware Attention-Enhanced DeepLabV3+ for Landslide Segmentation Along the Sichuan–Tibet Transportation Corridor
Accurate delineation of landslides from high-resolution optical imagery is important for hazard assessment and infrastructure safety in mountainous regions. However, automatic segmentation remains challenging because landslides vary greatly in scale, morphology, surface appearance, and boundary clarity. These difficulties are particularly evident for old landslides, whose surface characteristics may become less distinct after long-term erosion, vegetation recovery, and surface modification. To establish a consistent baseline, five representative semantic segmentation models were evaluated on the Landslide Recognition along the Sichuan–Tibet Transportation Corridor (LRSTTC) dataset under a unified experimental protocol. DeepLabV3+ showed the strongest overall baseline performance and was selected as the base architecture. We then developed a Boundary-Aware Attention-Enhanced DeepLabV3+ (BAA-DeepLabV3+) framework, in which a Convolutional Block Attention Module (CBAM) is applied after Atrous Spatial Pyramid Pooling (ASPP) for high-level feature refinement, while an auxiliary boundary branch provides additional spatial supervision during training. In the controlled ablation experiment with seed 42, BAA-DeepLabV3+ improved the Intersection over Union (IoU) from 0.3644 to 0.4229 and the Dice coefficient from 0.5341 to 0.5944, with only a negligible increase in model parameters. Across three random seeds, the mean IoU increased from 0.3599 ± 0.0275 to 0.3876 ± 0.0315. External validation on the independent Bijie landslide dataset further improved the mean IoU from 0.7295 to 0.7444, with BAA-DeepLabV3+ outperforming the baseline under all three random seeds. These results indicate that attention-based feature refinement and boundary-aware supervision can work complementarily to improve landslide segmentation while maintaining a favorable balance between segmentation performance and computational efficiency.
- Open Access
- 19 September 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (18) , 9283
CI-DIOR-7: A Task-Oriented Benchmark and Failure Analysis for Critical Infrastructure Detection in Remote Sensing Imagery
Accurate detection of critical infrastructure in high-resolution remote sensing imagery is important for infrastructure inventory, regional monitoring, and risk assessment. However, differences in object scale, spatial density, geometric form, and background complexity make it difficult to evaluate detectors using a single aggregate metric. This study develops a diagnostic evaluation framework utilizing CI-DIOR-7, a task-oriented benchmark derived from the DIOR dataset by retaining horizontal bounding-box annotations for seven infrastructure categories: Airport, Harbor, Bridge, Dam, Train Station, Storage Tank, and Windmill. The benchmark contains 23,463 images and 44,579 instances and preserves the original training, validation, and test splits. Four representative detectors—YOLO11s, RT-DETR-L, Faster R-CNN R50-FPN v2, and RetinaNet R50-FPN v2—were evaluated under a unified COCO-style evaluation protocol, modified with a maximum detection limit of 500 (maxDets = 500) to accurately evaluate highly dense infrastructure scenes. On the official test set of 11,738 images and 33,766 instances, RT-DETR-L achieved the highest mAP@0.5:0.95 of 0.3212 and the smallest validation-to-test performance drop. YOLO11s obtained an mAP@0.5:0.95 of 0.2785 while providing the lowest inference latency and the lowest false-positive burden on zero-GT images, at 0.168 false positives per image. Class-wise and difficulty-oriented analyses showed that Bridge was the most challenging category, with a mean class AP of 0.1514 and small-bridge recall ranging from 0.105 to 0.161 across the four models. Storage-tank detection was strongest in moderately dense scenes but deteriorated when more than 20 instances occurred in one image. These results show that no single architecture simultaneously achieved the highest detection accuracy, the lowest measured inference latency, and the lowest false-positive burden under the zero-GT protocol. CI-DIOR-7 therefore provides a reproducible benchmark and diagnostic framework for evaluating critical infrastructure detectors beyond overall mAP.
- 19 September 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (18) , 9283
Open AccessCI-DIOR-7: A Task-Oriented Benchmark and Failure Analysis for Critical Infrastructure Detection in Remote Sensing Imagery
Accurate detection of critical infrastructure in high-resolution remote sensing imagery is important for infrastructure inventory, regional monitoring, and risk assessment. However, differences in object scale, spatial density, geometric form, and background complexity make it difficult to evaluate detectors using a single aggregate metric. This study develops a diagnostic evaluation framework utilizing CI-DIOR-7, a task-oriented benchmark derived from the DIOR dataset by retaining horizontal bounding-box annotations for seven infrastructure categories: Airport, Harbor, Bridge, Dam, Train Station, Storage Tank, and Windmill. The benchmark contains 23,463 images and 44,579 instances and preserves the original training, validation, and test splits. Four representative detectors—YOLO11s, RT-DETR-L, Faster R-CNN R50-FPN v2, and RetinaNet R50-FPN v2—were evaluated under a unified COCO-style evaluation protocol, modified with a maximum detection limit of 500 (maxDets = 500) to accurately evaluate highly dense infrastructure scenes. On the official test set of 11,738 images and 33,766 instances, RT-DETR-L achieved the highest mAP@0.5:0.95 of 0.3212 and the smallest validation-to-test performance drop. YOLO11s obtained an mAP@0.5:0.95 of 0.2785 while providing the lowest inference latency and the lowest false-positive burden on zero-GT images, at 0.168 false positives per image. Class-wise and difficulty-oriented analyses showed that Bridge was the most challenging category, with a mean class AP of 0.1514 and small-bridge recall ranging from 0.105 to 0.161 across the four models. Storage-tank detection was strongest in moderately dense scenes but deteriorated when more than 20 instances occurred in one image. These results show that no single architecture simultaneously achieved the highest detection accuracy, the lowest measured inference latency, and the lowest false-positive burden under the zero-GT protocol. CI-DIOR-7 therefore provides a reproducible benchmark and diagnostic framework for evaluating critical infrastructure detectors beyond overall mAP.
- Open Access
- 3 March 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
A physics-guided machine learning framework for enhancing dust storm visibility prediction in arid and semi-arid regions
Dust sharply degrades visibility in arid and semi-arid regions, yet operational forecasting remains challenged by near-surface process errors in numerical weather prediction (NWP) and the poor generalization of purely data-driven models. We present a physics-guided machine learning (PGML) framework that post-processes European Centre for Medium-Range Weather Forecasts (ECMWF) forecasts to predict five ordinal visibility grades over the Kumtag Desert. A dust-lifecycle feature library (emission, vertical mixing, transport, wet scavenging) is coupled with an ordinal LightGBM architecture. On an independent test period, the model attains quadratic weighted kappa (QWK) of 0.26 (0–24 h), 0.17 (24–48 h), and 0.18 (48–72 h), with mean absolute error (MAE) of 0.48–0.56; gains versus data-only baselines increase with forecast horizon. Ablation experiments show that physics priors can effectively improve visibility prediction accuracy—reducing MAE by up to 10% and sustaining QWK beyond 24 h by constraining non-physical drift. Accordingly, the PGML visibility predictions show improved performance relative to data-only baselines. SHAP analysis reveals a forecast-horizon-dependent mechanistic shift from emission/surface-layer dynamics to stability-controlled vertical mixing, consistent with dust dynamics. The framework offers an interpretable, transferable paradigm for physics-constrained environmental forecasting.
- 3 March 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessA physics-guided machine learning framework for enhancing dust storm visibility prediction in arid and semi-arid regions
Dust sharply degrades visibility in arid and semi-arid regions, yet operational forecasting remains challenged by near-surface process errors in numerical weather prediction (NWP) and the poor generalization of purely data-driven models. We present a physics-guided machine learning (PGML) framework that post-processes European Centre for Medium-Range Weather Forecasts (ECMWF) forecasts to predict five ordinal visibility grades over the Kumtag Desert. A dust-lifecycle feature library (emission, vertical mixing, transport, wet scavenging) is coupled with an ordinal LightGBM architecture. On an independent test period, the model attains quadratic weighted kappa (QWK) of 0.26 (0–24 h), 0.17 (24–48 h), and 0.18 (48–72 h), with mean absolute error (MAE) of 0.48–0.56; gains versus data-only baselines increase with forecast horizon. Ablation experiments show that physics priors can effectively improve visibility prediction accuracy—reducing MAE by up to 10% and sustaining QWK beyond 24 h by constraining non-physical drift. Accordingly, the PGML visibility predictions show improved performance relative to data-only baselines. SHAP analysis reveals a forecast-horizon-dependent mechanistic shift from emission/surface-layer dynamics to stability-controlled vertical mixing, consistent with dust dynamics. The framework offers an interpretable, transferable paradigm for physics-constrained environmental forecasting.
- Open Access
Analysis of the Trends and Driving Factors of Cultivated Land Utilization Efficiency in Henan Province from 2000 to 2020
Amid persistent global food security challenges, the efficient utilization of cultivated land resources has become increasingly critical, as optimizing Cultivated Land Utilization Efficiency (CLUE) is paramount to ensuring food supply. This study introduced a cultivated land utilization index (CLUI) based on Fractional Vegetation Cover (FVC) to assess the spatiotemporal variations in Henan Province’s CLUE. The Theil–Sen slope and the Mann–Kendall test were used to analyze the spatiotemporal variations of CLUE in Henan Province from 2000 to 2020. Additionally, we used a genetic algorithm optimized Artificial Neural Network (ANN) and a particle swarm optimization-based Random Forest (RF) model to assess the comprehensive in-fluence between topography, climate, and human activities on CLUE, in which incorporating Shapley Additive Explanations (SHAP) values. The results reveal the following: (1) From 2000 to 2020, the CLUE in Henan province showed an overall upward trend, with strong spatial heterogeneity across various regions: the central and eastern areas generally showed decline, the northern region remained stable with slight increases, the western region saw significant growth, while the southern area exhibited complex fluctuations. (2) Natural and economic factors had notable impacts on CLUE in Henan province. Among these factors, population and economic factors played a dominant role, whereas average temperature exerted an inhibitory effect on CLUE in most parts of the province. (3) The influenced factors on CLUE varied spatially, with human activity impacts being more concentrated, while topographical and climatic influences were relatively dispersed. These findings provide a scientific basis for land management and agricultural policy formulation in major grain-producing areas, offering valuable insights into enhancing regional CLUE and promoting sustainable agricultural development.
Open AccessAnalysis of the Trends and Driving Factors of Cultivated Land Utilization Efficiency in Henan Province from 2000 to 2020
Amid persistent global food security challenges, the efficient utilization of cultivated land resources has become increasingly critical, as optimizing Cultivated Land Utilization Efficiency (CLUE) is paramount to ensuring food supply. This study introduced a cultivated land utilization index (CLUI) based on Fractional Vegetation Cover (FVC) to assess the spatiotemporal variations in Henan Province’s CLUE. The Theil–Sen slope and the Mann–Kendall test were used to analyze the spatiotemporal variations of CLUE in Henan Province from 2000 to 2020. Additionally, we used a genetic algorithm optimized Artificial Neural Network (ANN) and a particle swarm optimization-based Random Forest (RF) model to assess the comprehensive in-fluence between topography, climate, and human activities on CLUE, in which incorporating Shapley Additive Explanations (SHAP) values. The results reveal the following: (1) From 2000 to 2020, the CLUE in Henan province showed an overall upward trend, with strong spatial heterogeneity across various regions: the central and eastern areas generally showed decline, the northern region remained stable with slight increases, the western region saw significant growth, while the southern area exhibited complex fluctuations. (2) Natural and economic factors had notable impacts on CLUE in Henan province. Among these factors, population and economic factors played a dominant role, whereas average temperature exerted an inhibitory effect on CLUE in most parts of the province. (3) The influenced factors on CLUE varied spatially, with human activity impacts being more concentrated, while topographical and climatic influences were relatively dispersed. These findings provide a scientific basis for land management and agricultural policy formulation in major grain-producing areas, offering valuable insights into enhancing regional CLUE and promoting sustainable agricultural development.
- Open Access
ACO-TSSCD: An Optimized Deep Multimodal Temporal Semantic Segmentation Change Detection Approach for Monitoring Agricultural Land Conversion
With the acceleration of urbanization in agricultural areas and the continuous changes in land-use patterns, the transformation of agricultural land presents complexity and dynamism, which puts higher demands on precise monitoring. And most existing monitoring methods are constrained by limited spatial and temporal resolution, high computational demands, and challenges in distinguishing complex land cover types. These limitations hinder their ability to effectively detect rapid and subtle land use changes, particularly in areas experiencing rapid urban expansion, where their shortcomings become more pronounced. To address these challenges, this study presents a multimodal deep learning framework using a temporal semantic segmentation change detection (TSSCD) model optimized with ant colony optimization (ACO) to detect and analyze agricultural land conversion in Zhengzhou City, a major grain-producing area in China. This model utilizes Landsat 7/8 imagery and Sentinel-2 satellite imagery from 2003 to 2023 to capture the spatiotemporal transformation of cropland driven by urban expansion, infrastructure development, and population changes over the last two decades. The optimized TSSCD model achieves superior classification accuracy, with the kappa coefficient improving from 0.871 to 0.892, spatial F1 score from 0.903 to 0.935, and temporal F1 score from 0.848 to 0.879, indicating its effectiveness in identifying complex land-use changes. The significant spatiotemporal variation characteristics of agricultural land conversion in Zhengzhou City from 2003 to 2023 were revealed through the TSSCD model, with transformations initially concentrated near Zhengzhou’s urban core and expanding outward, particularly to the east and north. These results highlight the effectiveness of remote sensing and deep learning techniques in monitoring agricultural land conversion.
Open AccessACO-TSSCD: An Optimized Deep Multimodal Temporal Semantic Segmentation Change Detection Approach for Monitoring Agricultural Land Conversion
With the acceleration of urbanization in agricultural areas and the continuous changes in land-use patterns, the transformation of agricultural land presents complexity and dynamism, which puts higher demands on precise monitoring. And most existing monitoring methods are constrained by limited spatial and temporal resolution, high computational demands, and challenges in distinguishing complex land cover types. These limitations hinder their ability to effectively detect rapid and subtle land use changes, particularly in areas experiencing rapid urban expansion, where their shortcomings become more pronounced. To address these challenges, this study presents a multimodal deep learning framework using a temporal semantic segmentation change detection (TSSCD) model optimized with ant colony optimization (ACO) to detect and analyze agricultural land conversion in Zhengzhou City, a major grain-producing area in China. This model utilizes Landsat 7/8 imagery and Sentinel-2 satellite imagery from 2003 to 2023 to capture the spatiotemporal transformation of cropland driven by urban expansion, infrastructure development, and population changes over the last two decades. The optimized TSSCD model achieves superior classification accuracy, with the kappa coefficient improving from 0.871 to 0.892, spatial F1 score from 0.903 to 0.935, and temporal F1 score from 0.848 to 0.879, indicating its effectiveness in identifying complex land-use changes. The significant spatiotemporal variation characteristics of agricultural land conversion in Zhengzhou City from 2003 to 2023 were revealed through the TSSCD model, with transformations initially concentrated near Zhengzhou’s urban core and expanding outward, particularly to the east and north. These results highlight the effectiveness of remote sensing and deep learning techniques in monitoring agricultural land conversion.