Yong Ma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yong Ma has a total of 29 co-authors. Here are the top co-authors by number of shared publications:
- Erping Shang
Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
14 shared publications - Wutao Yao
Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
13 shared publications - Jin Yang
Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
9 shared publications - Shuyan Zhang
Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
8 shared publications - Fu Chen
Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
5 shared publications - Lixi Liu
Key Laboratory of Earth Observation of Hainan Province, Hainan Aerospace Information Research Institute, Sanya 572029, China
4 shared publications
Latest Publications
- Open Access
SRCT: Structure-Preserving Method for Sub-Meter Remote Sensing Image Super-Resolution
To address the scarcity of sub-meter remote sensing samples and structural inconsistencies such as edge blur and contour distortion in super-resolution reconstruction, this paper proposes SRCT, a super-resolution method tailored for sub-meter remote sensing imagery. The method consists of two parts: external structure guidance and internal structure optimization. External structure guidance is jointly realized by the structure encoder (SE) and structure guidance module (SGM): the SE extracts key structural features from high-resolution images, and the SGM injects these structural features into the super-resolution network layer by layer, achieving structural transfer from external priors to the reconstruction network. Internal structure optimization is handled by the backbone network SGCT, which introduces a dual-branch residual dense group (DBRDG): one branch uses window-based multi-head self-attention to model global geometric structures, and the other branch uses lightweight convolutions to model local texture features, enabling the network to adaptively balance structure and texture reconstruction internally. Experimental results show that SRCT clearly outperforms existing methods on structure-related metrics, with DISTS reduced by 8.7% and LPIPS reduced by 7.2%, and significantly improves reconstruction quality in structure-sensitive regions such as building contours and road continuity, providing a new technical route for sub-meter remote sensing image super-resolution reconstruction.
Open AccessSRCT: Structure-Preserving Method for Sub-Meter Remote Sensing Image Super-Resolution
To address the scarcity of sub-meter remote sensing samples and structural inconsistencies such as edge blur and contour distortion in super-resolution reconstruction, this paper proposes SRCT, a super-resolution method tailored for sub-meter remote sensing imagery. The method consists of two parts: external structure guidance and internal structure optimization. External structure guidance is jointly realized by the structure encoder (SE) and structure guidance module (SGM): the SE extracts key structural features from high-resolution images, and the SGM injects these structural features into the super-resolution network layer by layer, achieving structural transfer from external priors to the reconstruction network. Internal structure optimization is handled by the backbone network SGCT, which introduces a dual-branch residual dense group (DBRDG): one branch uses window-based multi-head self-attention to model global geometric structures, and the other branch uses lightweight convolutions to model local texture features, enabling the network to adaptively balance structure and texture reconstruction internally. Experimental results show that SRCT clearly outperforms existing methods on structure-related metrics, with DISTS reduced by 8.7% and LPIPS reduced by 7.2%, and significantly improves reconstruction quality in structure-sensitive regions such as building contours and road continuity, providing a new technical route for sub-meter remote sensing image super-resolution reconstruction.
- Open Access
Assessment of Suitable Habitats and Identification of Key Protection Areas for Polyplectron katsumatae in Jianfengling, Hainan Province, China
Polyplectron katsumatae is a rare and endangered species endemic to Hainan, China. It has long been regarded as a subspecies of the widely distributed Grey Peacock-Pheasant (Polyplectron bicalcaratum), a classification that has resulted in a paucity of targeted conservation studies and rendered efforts to protect and restore its populations and habitats exceedingly challenging. In this study, the Jianfengling section of Hainan Tropical Rainforest National Park was designated as the research area. We comprehensively utilized infrared camera monitoring data for P. katsumatae and other species, alongside habitat environmental factor data obtained through multiple monitoring approaches. An ensemble species distribution model (ESDM) was employed to evaluate the habitat suitability for four ground-dwelling bird species, including P. katsumatae, and to investigate their environmental preferences and competitive interactions during habitat selection. Subsequently, the Marxan model was applied to identify key protection areas for P. katsumatae. The results indicate that the suitable habitat for P. katsumatae is primarily distributed in the central, eastern, and certain southern areas of the study region, with low spatial overlap and minimal competition from the suitable habitats of the other three ground-dwelling bird species. However, due to anthropogenic disturbances and the inherently stringent habitat requirements of P. katsumatae, its overall suitable habitat area is limited, exhibiting a concentrated distribution overall with fragmented, small patches within it. Our study recommends designating the eastern and southern regions of the study area as key protection areas for P. katsumatae, thereby providing a robust baseline environment and policy support for the targeted protection of its habitat and the recovery of its populations.
Open AccessAssessment of Suitable Habitats and Identification of Key Protection Areas for Polyplectron katsumatae in Jianfengling, Hainan Province, China
Polyplectron katsumatae is a rare and endangered species endemic to Hainan, China. It has long been regarded as a subspecies of the widely distributed Grey Peacock-Pheasant (Polyplectron bicalcaratum), a classification that has resulted in a paucity of targeted conservation studies and rendered efforts to protect and restore its populations and habitats exceedingly challenging. In this study, the Jianfengling section of Hainan Tropical Rainforest National Park was designated as the research area. We comprehensively utilized infrared camera monitoring data for P. katsumatae and other species, alongside habitat environmental factor data obtained through multiple monitoring approaches. An ensemble species distribution model (ESDM) was employed to evaluate the habitat suitability for four ground-dwelling bird species, including P. katsumatae, and to investigate their environmental preferences and competitive interactions during habitat selection. Subsequently, the Marxan model was applied to identify key protection areas for P. katsumatae. The results indicate that the suitable habitat for P. katsumatae is primarily distributed in the central, eastern, and certain southern areas of the study region, with low spatial overlap and minimal competition from the suitable habitats of the other three ground-dwelling bird species. However, due to anthropogenic disturbances and the inherently stringent habitat requirements of P. katsumatae, its overall suitable habitat area is limited, exhibiting a concentrated distribution overall with fragmented, small patches within it. Our study recommends designating the eastern and southern regions of the study area as key protection areas for P. katsumatae, thereby providing a robust baseline environment and policy support for the targeted protection of its habitat and the recovery of its populations.
- Open Access
- 31 March 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15  (7) , 3817
Dynamics of Cadmium Pollution Risk in Agricultural Land Soil of Tropical Islands in China from 2000 to 2024: A Case Study of Hainan Island
The issue of cadmium (Cd) contamination in agricultural soils has garnered significant attention due to its implications for food safety and environmental health. However, research on its spatial and temporal distribution in tropical agricultural zones with high intensity cropping remains limited. This study conducts a meta-analysis of 38 studies published between 2000 and 2024 to comprehensively assess Cd pollution on Hainan Island, China, a tropical region with three cropping cycles per year. By synthesizing 165 records from 4110 sampling sites, the analysis reveals that the mean Cd concentration stands at 0.12 mg/kg, with peak values reaching 3.11 mg/kg. Approximately 80.44% of sites fall into the categories of unpolluted or unpolluted to moderately unpolluted based on the geo-accumulation index, while 2.17% of sites exhibited heavy to extreme Cd accumulation, and 2.13% posed an high ecological risk (level IV). Spatially, Cd concentrations were lower in central areas, such as Baisha County, but higher in the northwest and southeast, notably in Haikou City in the northwest and Sanya City in the south. Over time, Cd pollution has intensified, with mean concentrations rising from 0.09 mg/kg (2005–2014) to 0.16 mg/kg (2017–2023), with exceedance rates rising from around 1% to between 7.78% and 20.93%, peaking in 2017. The increasing trend of cadmium accumulation is closely related to human activities, with agricultural, industrial, and transportation-related activities playing an increasingly significant role in Cd accumulation. Pesticide use, phosphate fertilizer application, and industrial emissions have been identified as major contributing factors. These findings highlight the need for targeted mitigation strategies to control pollution sources and ensure the long-term sustainability of agricultural production on Hainan Island.
- 31 March 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15  (7) , 3817
Open AccessDynamics of Cadmium Pollution Risk in Agricultural Land Soil of Tropical Islands in China from 2000 to 2024: A Case Study of Hainan Island
The issue of cadmium (Cd) contamination in agricultural soils has garnered significant attention due to its implications for food safety and environmental health. However, research on its spatial and temporal distribution in tropical agricultural zones with high intensity cropping remains limited. This study conducts a meta-analysis of 38 studies published between 2000 and 2024 to comprehensively assess Cd pollution on Hainan Island, China, a tropical region with three cropping cycles per year. By synthesizing 165 records from 4110 sampling sites, the analysis reveals that the mean Cd concentration stands at 0.12 mg/kg, with peak values reaching 3.11 mg/kg. Approximately 80.44% of sites fall into the categories of unpolluted or unpolluted to moderately unpolluted based on the geo-accumulation index, while 2.17% of sites exhibited heavy to extreme Cd accumulation, and 2.13% posed an high ecological risk (level IV). Spatially, Cd concentrations were lower in central areas, such as Baisha County, but higher in the northwest and southeast, notably in Haikou City in the northwest and Sanya City in the south. Over time, Cd pollution has intensified, with mean concentrations rising from 0.09 mg/kg (2005–2014) to 0.16 mg/kg (2017–2023), with exceedance rates rising from around 1% to between 7.78% and 20.93%, peaking in 2017. The increasing trend of cadmium accumulation is closely related to human activities, with agricultural, industrial, and transportation-related activities playing an increasingly significant role in Cd accumulation. Pesticide use, phosphate fertilizer application, and industrial emissions have been identified as major contributing factors. These findings highlight the need for targeted mitigation strategies to control pollution sources and ensure the long-term sustainability of agricultural production on Hainan Island.
- Open Access
- 9 March 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17  (6) , 963
MCACD: A Multi-Scale Convolutional Attention Network for Forest and Grassland Change Detection
Forests and grasslands are crucial components of terrestrial ecosystems. Their stability and health play an essential role in maintaining ecological balance. However, with the intensification of human activities, these ecosystems are facing increasing risks of degradation. Forest and grassland change detection is vital for monitoring and management by governmental agencies, requiring reliable and automated detection models. Currently, most deep learning-based change detection models face significant limitations in achieving precise detection of forest and grassland changes. These challenges are particularly evident in addressing incomplete segmentation objects and accurately discriminating hard/challenging samples, highlighting substantial room for further research and improvement. To address these challenges, we have developed MCACD. This model effectively preserves detailed information by combining high-resolution feature maps with dynamic upsampling segmentation heads. It also integrates an efficient feature fusion module, designed by us, to enable the integration of dual-temporal features. Additionally, we introduce a highly efficient multi-scale cascade fully convolutional attention decoder, which progressively enhances spatial features at multiple scales and resolutions. A deep supervision strategy within the decoder further optimizes the feature learning process. Experimental results show that, compared to other advanced change detection methods, MCACD performs exceptionally well on our custom-built forest and grassland change detection dataset. It achieves the best F1 score and IoU values, demonstrating significant potential for practical applications in forest and grassland change detection.
- 9 March 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17  (6) , 963
Open AccessMCACD: A Multi-Scale Convolutional Attention Network for Forest and Grassland Change Detection
Forests and grasslands are crucial components of terrestrial ecosystems. Their stability and health play an essential role in maintaining ecological balance. However, with the intensification of human activities, these ecosystems are facing increasing risks of degradation. Forest and grassland change detection is vital for monitoring and management by governmental agencies, requiring reliable and automated detection models. Currently, most deep learning-based change detection models face significant limitations in achieving precise detection of forest and grassland changes. These challenges are particularly evident in addressing incomplete segmentation objects and accurately discriminating hard/challenging samples, highlighting substantial room for further research and improvement. To address these challenges, we have developed MCACD. This model effectively preserves detailed information by combining high-resolution feature maps with dynamic upsampling segmentation heads. It also integrates an efficient feature fusion module, designed by us, to enable the integration of dual-temporal features. Additionally, we introduce a highly efficient multi-scale cascade fully convolutional attention decoder, which progressively enhances spatial features at multiple scales and resolutions. A deep supervision strategy within the decoder further optimizes the feature learning process. Experimental results show that, compared to other advanced change detection methods, MCACD performs exceptionally well on our custom-built forest and grassland change detection dataset. It achieves the best F1 score and IoU values, demonstrating significant potential for practical applications in forest and grassland change detection.
- Open Access
Habitat Suitability Assessment of Key Wildlife in Hainan Tropical Rainforest Based on ESDM
Hainan tropical rainforest is the largest contiguous tropical rainforest in China, but it has experienced increasing disturbances from anthropogenic activities in recent decades due to economic and social development. However, the current status of wildlife habitats within the rainforest remains insufficiently studied, lacking systematic and scientific assessments necessary to guide effective biodiversity conservation strategies. This study focuses on Jianfengling area of Hainan tropical rainforest, using wildlife infrared camera monitoring data and habitat environmental factor data collected through multi-source monitoring in 2020–2021. By applying the Ensemble Species Distribution Model (ESDM), we assessed the spatial distribution of habitat suitability and its influencing factors for seven representative wildlife species, as well as the overall spatial distribution of multi-species habitat suitability. The results indicate that wildlife habitat suitability in Jianfengling study area exhibits a spatial pattern of high suitability in the central regions and low suitability in surrounding areas. Anthropogenic activities and DEM were identified as the most significant factors influencing habitat selection, with most species favoring mid and high altitude areas (500–1000 m) where human activities are less prevalent. This study provides scientific support for tropical rainforest management authorities to optimize resource allocation, develop dynamic monitoring strategies, and implement effective conservation measures.
Open AccessHabitat Suitability Assessment of Key Wildlife in Hainan Tropical Rainforest Based on ESDM
Hainan tropical rainforest is the largest contiguous tropical rainforest in China, but it has experienced increasing disturbances from anthropogenic activities in recent decades due to economic and social development. However, the current status of wildlife habitats within the rainforest remains insufficiently studied, lacking systematic and scientific assessments necessary to guide effective biodiversity conservation strategies. This study focuses on Jianfengling area of Hainan tropical rainforest, using wildlife infrared camera monitoring data and habitat environmental factor data collected through multi-source monitoring in 2020–2021. By applying the Ensemble Species Distribution Model (ESDM), we assessed the spatial distribution of habitat suitability and its influencing factors for seven representative wildlife species, as well as the overall spatial distribution of multi-species habitat suitability. The results indicate that wildlife habitat suitability in Jianfengling study area exhibits a spatial pattern of high suitability in the central regions and low suitability in surrounding areas. Anthropogenic activities and DEM were identified as the most significant factors influencing habitat selection, with most species favoring mid and high altitude areas (500–1000 m) where human activities are less prevalent. This study provides scientific support for tropical rainforest management authorities to optimize resource allocation, develop dynamic monitoring strategies, and implement effective conservation measures.
- Open Access
- 27 December 2024
- Published by MDPI AG in Remote Sensing
- Vol. 17  (1) , 46
Spatial and Temporal Changes and Assessment of Multi-Species Habitat in Hainan Jianfengling Protected Area
The loss and fragmentation of wildlife habitats is a major threat to their survival and expansion, and protected areas (PAs) are the main tool for conserving biodiversity and protecting habitats. However, most current studies focus on analyzing suitable habitats for species and rarely analyze the spatial and temporal changes in multi-species habitats in protected areas and the effectiveness of conservation. In this study, we analyzed changes in the suitable habitats of five focal mammal species before and after the incorporation of the Hainan Jianfengling protected area into China’s national parks. We utilized the ensemble species distribution model (ESDM) to assess these changes, based on multi-species infrared camera monitoring data from 2015 to 2016 and 2020 to 2021. Furthermore, we evaluated differences in conservation effectiveness before and after the establishment of the national parks. The results showed that there were some differences in habitat changes among all the species included in this study, and all of them showed the phenomenon of the migration from suitable habitats to the central area. The environmental changes in and around the protected area suggest that the changes are closely related to the increase in anthropogenic activities around the protected area, and it is recommended that the protected area should be better managed at its edges to minimize the impact of anthropogenic disturbances on the species and their habitats.
- 27 December 2024
- Published by MDPI AG in Remote Sensing
- Vol. 17  (1) , 46
Open AccessSpatial and Temporal Changes and Assessment of Multi-Species Habitat in Hainan Jianfengling Protected Area
The loss and fragmentation of wildlife habitats is a major threat to their survival and expansion, and protected areas (PAs) are the main tool for conserving biodiversity and protecting habitats. However, most current studies focus on analyzing suitable habitats for species and rarely analyze the spatial and temporal changes in multi-species habitats in protected areas and the effectiveness of conservation. In this study, we analyzed changes in the suitable habitats of five focal mammal species before and after the incorporation of the Hainan Jianfengling protected area into China’s national parks. We utilized the ensemble species distribution model (ESDM) to assess these changes, based on multi-species infrared camera monitoring data from 2015 to 2016 and 2020 to 2021. Furthermore, we evaluated differences in conservation effectiveness before and after the establishment of the national parks. The results showed that there were some differences in habitat changes among all the species included in this study, and all of them showed the phenomenon of the migration from suitable habitats to the central area. The environmental changes in and around the protected area suggest that the changes are closely related to the increase in anthropogenic activities around the protected area, and it is recommended that the protected area should be better managed at its edges to minimize the impact of anthropogenic disturbances on the species and their habitats.
- Open Access
Trends in Heavy Metal Pollution in Agricultural Land Soils of Tropical Islands in China (2000–2024): A Case Study on Hainan Island
Heavy metal contamination in agricultural soils has garnered increasing attention, yet research on the spatiotemporal trends of heavy metal pollution in tropical regions with multiple annual crop harvests remains limited. This study examines data from 41 studies published between 2000 and 2024, including 206 records from 4122 sampling points on Hainan Island in China, to investigate the spatial distribution and temporal trends of heavy metal pollution. The results reveal that the average concentrations of Cd, Pb, As, Cr, and Hg in surface soil samples from agricultural lands on Hainan Island are 0.12, 28.28, 4.36, 63.98, and 0.075 mg/kg, respectively, all below the risk screening thresholds set by the Soil Pollution Risk Control Standard for Agricultural Land (GB 15618-2018). Spatially, heavy metal concentrations exhibit considerable regional variation. Cd levels are lower in the central region but higher in the northern and southern parts of the island. Both the cumulative pollution index and potential ecological risk index are elevated at the northern and southern ends, indicating more severe pollution in these areas. Pb and As show similar spatial patterns, with higher concentrations in the west and lower concentrations in the east. Conversely, Cr has higher concentrations in the northeast and lower concentrations in the southwest. Hg levels are elevated at the northern and southern ends of the island, though the overall pollution and ecological risk in these areas remain relatively low. Temporally, the concentration of heavy metals in agricultural soils has increased overall over the past two decades, with peak values occurring between 2017 and 2023. From 2002 to 2013, the variation was modest, while the largest fluctuations occurred between 2014 and 2016. Among the metals, Cr exhibited the most significant increase, indicating the most severe pollution, followed by Cd and Hg. As and Pb showed relatively lower levels of contamination. Regarding exceedance rates, the exceedances were evaluated against the thresholds established in GB15618-2018 and GB15618-1995. Cd’s exceedance rate increased from approximately 1% between 2002 and 2014 to between 7.78% and 20.93% in the following years, peaking in 2017. The exceedance rate for As rose slightly from 0% to 0.83%, with sporadic exceedances starting in 2015. Although these were relatively minor, a severe pollution point for As was observed in 2019. Exceedance rates for Pb and Cr increased significantly, from 0.75% and 7.50% in 2019 to 1.94% and 9.44% in 2023, reflecting increases of 4.8 to 10 times. These findings underscore the need for strengthened monitoring and management of heavy metal pollution in agricultural soils on Hainan Island to safeguard land quality and ensure the sustainability of local agricultural practices.
Open AccessTrends in Heavy Metal Pollution in Agricultural Land Soils of Tropical Islands in China (2000–2024): A Case Study on Hainan Island
Heavy metal contamination in agricultural soils has garnered increasing attention, yet research on the spatiotemporal trends of heavy metal pollution in tropical regions with multiple annual crop harvests remains limited. This study examines data from 41 studies published between 2000 and 2024, including 206 records from 4122 sampling points on Hainan Island in China, to investigate the spatial distribution and temporal trends of heavy metal pollution. The results reveal that the average concentrations of Cd, Pb, As, Cr, and Hg in surface soil samples from agricultural lands on Hainan Island are 0.12, 28.28, 4.36, 63.98, and 0.075 mg/kg, respectively, all below the risk screening thresholds set by the Soil Pollution Risk Control Standard for Agricultural Land (GB 15618-2018). Spatially, heavy metal concentrations exhibit considerable regional variation. Cd levels are lower in the central region but higher in the northern and southern parts of the island. Both the cumulative pollution index and potential ecological risk index are elevated at the northern and southern ends, indicating more severe pollution in these areas. Pb and As show similar spatial patterns, with higher concentrations in the west and lower concentrations in the east. Conversely, Cr has higher concentrations in the northeast and lower concentrations in the southwest. Hg levels are elevated at the northern and southern ends of the island, though the overall pollution and ecological risk in these areas remain relatively low. Temporally, the concentration of heavy metals in agricultural soils has increased overall over the past two decades, with peak values occurring between 2017 and 2023. From 2002 to 2013, the variation was modest, while the largest fluctuations occurred between 2014 and 2016. Among the metals, Cr exhibited the most significant increase, indicating the most severe pollution, followed by Cd and Hg. As and Pb showed relatively lower levels of contamination. Regarding exceedance rates, the exceedances were evaluated against the thresholds established in GB15618-2018 and GB15618-1995. Cd’s exceedance rate increased from approximately 1% between 2002 and 2014 to between 7.78% and 20.93% in the following years, peaking in 2017. The exceedance rate for As rose slightly from 0% to 0.83%, with sporadic exceedances starting in 2015. Although these were relatively minor, a severe pollution point for As was observed in 2019. Exceedance rates for Pb and Cr increased significantly, from 0.75% and 7.50% in 2019 to 1.94% and 9.44% in 2023, reflecting increases of 4.8 to 10 times. These findings underscore the need for strengthened monitoring and management of heavy metal pollution in agricultural soils on Hainan Island to safeguard land quality and ensure the sustainability of local agricultural practices.
- Open Access
- 16 August 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16  (16) , 3013
Production of Annual Nighttime Light Based on De-Difference Smoothing Algorithm
Nighttime light (NTL) remote sensing has emerged as a powerful tool in various fields such as urban expansion, socio-economic estimation, light pollution, and energy domains. However, current annual NTL products suffer from several critical limitations, including poor consistency, severe background noise, and limited comparability. These issues have significantly interfered with the research of long-term NTL trends and diminished the accuracy of related findings. Therefore, this study developed a de-difference smoothing algorithm for producing high-quality annual NTL products based on monthly National Polar-orbiting Partnership Visible Infrared Imaging Radiometer Suite (NPP-VIIRS) NTL data. It enabled the construction of a continuous global high-quality NTL dataset, named the De-Difference Smoothed Nighttime Light (DDSNL), covering the period from 2012 to 2023. Comparative analyses were conducted to validate the accuracy and availability of the DDSNL product against the benchmark EOG NPP-VIIRS and NPP-VIIRS-like NTL datasets. The results showed that DDSNL products had strong correlation with the NTL distribution of EOG NPP-VIIRS, but little correlation with NPP-VIIRS-like. Notably, DDSNL demonstrated better background noise reduction and higher separability between NTL and non-NTL areas compared to EOG NPP-VIIRS NTL. In contrast to the complete exclusion of background in NPP-VIIRS-Like, the retention of background values in DDSNL leads to more reasonable representation in the urban fringes. In the analysis of NTL changes matching impervious surface changes, the DDSNL product demonstrated the least interference from noise, resulting in the smallest segmentation threshold and the highest matching accuracy. This indirectly demonstrates the spatial and temporal consistency of the annual DDSNL product, ensuring its reliability in change detection-related studies. The annual DDSNL product developed in this research exhibits high fidelity, strong consistency, and improved comparability, and can provide reliable data reference for applications in electrification and urban studies.
- 16 August 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16  (16) , 3013
Open AccessProduction of Annual Nighttime Light Based on De-Difference Smoothing Algorithm
Nighttime light (NTL) remote sensing has emerged as a powerful tool in various fields such as urban expansion, socio-economic estimation, light pollution, and energy domains. However, current annual NTL products suffer from several critical limitations, including poor consistency, severe background noise, and limited comparability. These issues have significantly interfered with the research of long-term NTL trends and diminished the accuracy of related findings. Therefore, this study developed a de-difference smoothing algorithm for producing high-quality annual NTL products based on monthly National Polar-orbiting Partnership Visible Infrared Imaging Radiometer Suite (NPP-VIIRS) NTL data. It enabled the construction of a continuous global high-quality NTL dataset, named the De-Difference Smoothed Nighttime Light (DDSNL), covering the period from 2012 to 2023. Comparative analyses were conducted to validate the accuracy and availability of the DDSNL product against the benchmark EOG NPP-VIIRS and NPP-VIIRS-like NTL datasets. The results showed that DDSNL products had strong correlation with the NTL distribution of EOG NPP-VIIRS, but little correlation with NPP-VIIRS-like. Notably, DDSNL demonstrated better background noise reduction and higher separability between NTL and non-NTL areas compared to EOG NPP-VIIRS NTL. In contrast to the complete exclusion of background in NPP-VIIRS-Like, the retention of background values in DDSNL leads to more reasonable representation in the urban fringes. In the analysis of NTL changes matching impervious surface changes, the DDSNL product demonstrated the least interference from noise, resulting in the smallest segmentation threshold and the highest matching accuracy. This indirectly demonstrates the spatial and temporal consistency of the annual DDSNL product, ensuring its reliability in change detection-related studies. The annual DDSNL product developed in this research exhibits high fidelity, strong consistency, and improved comparability, and can provide reliable data reference for applications in electrification and urban studies.
- Open Access
- 5 March 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16  (5) , 913
YOLOv7oSAR: A Lightweight High-Precision Ship Detection Model for SAR Images Based on the YOLOv7 Algorithm
Researchers have explored various methods to fully exploit the all-weather characteristics of Synthetic aperture radar (SAR) images to achieve high-precision, real-time, computationally efficient, and easily deployable ship target detection models. These methods include Constant False Alarm Rate (CFAR) algorithms and deep learning approaches such as RCNN, YOLO, and SSD, among others. While these methods outperform traditional algorithms in SAR ship detection, challenges still exist in handling the arbitrary ship distributions and small target features in SAR remote sensing images. Existing models are complex, with a large number of parameters, hindering effective deployment. This paper introduces a YOLOv7 oriented bounding box SAR ship detection model (YOLOv7oSAR). The model employs a rotation box detection mechanism, uses the KLD loss function to enhance accuracy, and introduces a Bi-former attention mechanism to improve small target detection. By redesigning the network’s width and depth and incorporating a lightweight P-ELAN structure, the model effectively reduces its size and computational requirements. The proposed model achieves high-precision detection results on the public RSDD dataset (94.8% offshore, 66.6% nearshore), and its generalization ability is validated on a custom dataset (94.2% overall detection accuracy).
- 5 March 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16  (5) , 913
Open AccessYOLOv7oSAR: A Lightweight High-Precision Ship Detection Model for SAR Images Based on the YOLOv7 Algorithm
Researchers have explored various methods to fully exploit the all-weather characteristics of Synthetic aperture radar (SAR) images to achieve high-precision, real-time, computationally efficient, and easily deployable ship target detection models. These methods include Constant False Alarm Rate (CFAR) algorithms and deep learning approaches such as RCNN, YOLO, and SSD, among others. While these methods outperform traditional algorithms in SAR ship detection, challenges still exist in handling the arbitrary ship distributions and small target features in SAR remote sensing images. Existing models are complex, with a large number of parameters, hindering effective deployment. This paper introduces a YOLOv7 oriented bounding box SAR ship detection model (YOLOv7oSAR). The model employs a rotation box detection mechanism, uses the KLD loss function to enhance accuracy, and introduces a Bi-former attention mechanism to improve small target detection. By redesigning the network’s width and depth and incorporating a lightweight P-ELAN structure, the model effectively reduces its size and computational requirements. The proposed model achieves high-precision detection results on the public RSDD dataset (94.8% offshore, 66.6% nearshore), and its generalization ability is validated on a custom dataset (94.2% overall detection accuracy).
- Open Access
- 31 March 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15  (7) , 1863
Temporal Co-Attention Guided Conditional Generative Adversarial Network for Optical Image Synthesis
In the field of SAR-to-optical image synthesis, current methods based on conditional generative adversarial networks (CGANs) have satisfying performance under simple scenarios, but the performance drops severely under complicated scenarios. Considering that SAR images can form a robust time series due to SAR’s all-weather imaging ability, we take advantage of this and extract a temporal correlation from bi-temporal SAR images to guide the translation. To achieve this, we introduce a co-attention mechanism into the CGAN that learns the correlation between optically-available and optically-absent time points, selectively enhances the features of the former time point, and eventually guides the model to a better optical image synthesis on the latter time point. Additionally, we adopt a strategy to balance the weight of optical and SAR features to extract better features from the SAR input. With these strategies, the quality of synthesized images is notably improved in complicated scenarios. The synthesized images can increase the spatial and temporal resolution of optical imagery, greatly improving the availability of data for the applications of crop monitoring, change detection, and visual interpretation.
- 31 March 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15  (7) , 1863
Open AccessTemporal Co-Attention Guided Conditional Generative Adversarial Network for Optical Image Synthesis
In the field of SAR-to-optical image synthesis, current methods based on conditional generative adversarial networks (CGANs) have satisfying performance under simple scenarios, but the performance drops severely under complicated scenarios. Considering that SAR images can form a robust time series due to SAR’s all-weather imaging ability, we take advantage of this and extract a temporal correlation from bi-temporal SAR images to guide the translation. To achieve this, we introduce a co-attention mechanism into the CGAN that learns the correlation between optically-available and optically-absent time points, selectively enhances the features of the former time point, and eventually guides the model to a better optical image synthesis on the latter time point. Additionally, we adopt a strategy to balance the weight of optical and SAR features to extract better features from the SAR input. With these strategies, the quality of synthesized images is notably improved in complicated scenarios. The synthesized images can increase the spatial and temporal resolution of optical imagery, greatly improving the availability of data for the applications of crop monitoring, change detection, and visual interpretation.