Chaofan Pan, Runsheng Li, Q. Hu, C. Niu, Wei Liu, Wanjie Lu
Fine-grained classification of ship targets is an important task in remote sensing, having numerous applications in military reconnaissance and sea surveillance. Due to the influence of various imaging factors, ship targets in remote sensing images have considerable inter-class similarity and intra-class difference, which brings significant challenges to fine-grained classification. In response, we developed a contrastive learning network based on causal attention (C2Net) to improve the model’s fine-grained identification ability from local details. The asynchronous feature learning mode of “decoupling + aggregation” is adopted to reduce the mutual influence between local features and improve the quality of local features. In the decoupling stage, the feature vectors of each part of the ship targets are de-correlated using a decoupling function to prevent feature adhesion. Considering the possibility of false associations between results and features, the decoupled part is designed based on the counterfactual causal attention network to enhance the model’s predictive logic. In the aggregation stage, the local attention weight learned in the decoupling stage is used to carry out feature fusion on the trunk feature weight. Then, the proposed feature re-association module is used to re-associate and integrate the target local information contained in the fusion feature to obtain the target feature vector. Finally, the aggregation function is used to complete the clustering process of the target feature vectors and fine-grained classification is realized. Using two large-scale datasets, the experimental results show that the proposed C2Net method had better fine-grained classification than other methods.
{"title":"Contrastive Learning Network Based on Causal Attention for Fine-Grained Ship Classification in Remote Sensing Scenarios","authors":"Chaofan Pan, Runsheng Li, Q. Hu, C. Niu, Wei Liu, Wanjie Lu","doi":"10.3390/rs15133393","DOIUrl":"https://doi.org/10.3390/rs15133393","url":null,"abstract":"Fine-grained classification of ship targets is an important task in remote sensing, having numerous applications in military reconnaissance and sea surveillance. Due to the influence of various imaging factors, ship targets in remote sensing images have considerable inter-class similarity and intra-class difference, which brings significant challenges to fine-grained classification. In response, we developed a contrastive learning network based on causal attention (C2Net) to improve the model’s fine-grained identification ability from local details. The asynchronous feature learning mode of “decoupling + aggregation” is adopted to reduce the mutual influence between local features and improve the quality of local features. In the decoupling stage, the feature vectors of each part of the ship targets are de-correlated using a decoupling function to prevent feature adhesion. Considering the possibility of false associations between results and features, the decoupled part is designed based on the counterfactual causal attention network to enhance the model’s predictive logic. In the aggregation stage, the local attention weight learned in the decoupling stage is used to carry out feature fusion on the trunk feature weight. Then, the proposed feature re-association module is used to re-associate and integrate the target local information contained in the fusion feature to obtain the target feature vector. Finally, the aggregation function is used to complete the clustering process of the target feature vectors and fine-grained classification is realized. Using two large-scale datasets, the experimental results show that the proposed C2Net method had better fine-grained classification than other methods.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87597828","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hao Song, Mengya Sheng, L. Lei, Kaiyuan Guo, Shaoqing Zhang, Zhanghui Ji
Space-based measurements, such as the Greenhouse gases Observing SATellite (GOSAT) and the TROPOspheric Monitoring Instrument (TROPOMI) aboard the Sentinel-5 Precursor satellite, provide global observations of the column-averaged CH4 concentration (XCH4). Due to the irregular observations and data gaps in the retrievals, studies on the spatial and temporal variations of regional atmospheric CH4 concentrations are limited. In this paper, we mapped XCH4 data over monsoon Asia using GOSAT and TROPOMI observations from April 2009 to December 2021 and analyzed the spatial and temporal pattern of atmospheric CH4 variations and emissions. The results show that atmospheric CH4 concentrations over monsoon Asia have long-term increases with an annual growth rate of roughly 8.4 ppb. The spatial and temporal trends of XCH4 data are significantly correlated with anthropogenic CH4 emissions from the bottom-up emission inventory of EDGAR. The spatial pattern of gridded XCH4 temporal variations in China presents a basically consistent distribution with the Heihe–Tengchong Line, which is mainly related to the difference in anthropogenic emissions in the eastern and western areas. Using the mapping of XCH4 data from 2019 to 2021, this study further revealed the response of atmospheric CH4 concentrations to anthropogenic emissions in different urban agglomerations. For the urban agglomerations, the triangle of Central China (TCC), the Chengdu–Chongqing City Group (CCG), and the Yangtze River Delta (YRD) show higher CH4 concentrations and emissions than the Beijing–Tianjin–Hebei region and nearby areas (BTH). The results reveal the spatial and temporal distribution of CH4 concentrations and quantify the differences between urban agglomerations, which will support further studies on the drivers of methane emissions.
{"title":"Spatial and Temporal Variations of Atmospheric CH4 in Monsoon Asia Detected by Satellite Observations of GOSAT and TROPOMI","authors":"Hao Song, Mengya Sheng, L. Lei, Kaiyuan Guo, Shaoqing Zhang, Zhanghui Ji","doi":"10.3390/rs15133389","DOIUrl":"https://doi.org/10.3390/rs15133389","url":null,"abstract":"Space-based measurements, such as the Greenhouse gases Observing SATellite (GOSAT) and the TROPOspheric Monitoring Instrument (TROPOMI) aboard the Sentinel-5 Precursor satellite, provide global observations of the column-averaged CH4 concentration (XCH4). Due to the irregular observations and data gaps in the retrievals, studies on the spatial and temporal variations of regional atmospheric CH4 concentrations are limited. In this paper, we mapped XCH4 data over monsoon Asia using GOSAT and TROPOMI observations from April 2009 to December 2021 and analyzed the spatial and temporal pattern of atmospheric CH4 variations and emissions. The results show that atmospheric CH4 concentrations over monsoon Asia have long-term increases with an annual growth rate of roughly 8.4 ppb. The spatial and temporal trends of XCH4 data are significantly correlated with anthropogenic CH4 emissions from the bottom-up emission inventory of EDGAR. The spatial pattern of gridded XCH4 temporal variations in China presents a basically consistent distribution with the Heihe–Tengchong Line, which is mainly related to the difference in anthropogenic emissions in the eastern and western areas. Using the mapping of XCH4 data from 2019 to 2021, this study further revealed the response of atmospheric CH4 concentrations to anthropogenic emissions in different urban agglomerations. For the urban agglomerations, the triangle of Central China (TCC), the Chengdu–Chongqing City Group (CCG), and the Yangtze River Delta (YRD) show higher CH4 concentrations and emissions than the Beijing–Tianjin–Hebei region and nearby areas (BTH). The results reveal the spatial and temporal distribution of CH4 concentrations and quantify the differences between urban agglomerations, which will support further studies on the drivers of methane emissions.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83277709","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cambodia has the most fires per area in Southeast Asia, with fire activity have significantly increased since the early 2000s. Wildfire occurrences are multi-factorial in nature, and isolating the relative contribution of each driver remains a challenge. In this study, we quantify the relative importance of each driver of fire by analyzing annual spatial regression models of fire occurrence across Cambodia from 2003 to 2020. Our models demonstrated satisfactory performance, explaining 69 to 81% of the variance in fire occurrence. We found that deforestation was consistently the dominant driver of fire across 48 to 70% of the country throughout the study period. Although the influence of low precipitation on fires has increased in 2019 and 2020, the period is not long enough to establish any significant trends. During the study period, wind speed, elevation, and soil moisture had a slight influence of 6–20% without any clear trend, indicating that deforestation continues to be the main driver of fire. Our study improves the current understanding of the drivers of biomass fires across Cambodia, and the methodological framework developed here (quantitative decoupling of the drivers) has strong potential to be applied to other fire-prone areas around the world.
{"title":"Deforestation as the Prominent Driver of the Intensifying Wildfire in Cambodia, Revealed through Geospatial Analysis","authors":"Min-Sung Sim, S. Wee, Enner H. Alcântara, E. Park","doi":"10.3390/rs15133388","DOIUrl":"https://doi.org/10.3390/rs15133388","url":null,"abstract":"Cambodia has the most fires per area in Southeast Asia, with fire activity have significantly increased since the early 2000s. Wildfire occurrences are multi-factorial in nature, and isolating the relative contribution of each driver remains a challenge. In this study, we quantify the relative importance of each driver of fire by analyzing annual spatial regression models of fire occurrence across Cambodia from 2003 to 2020. Our models demonstrated satisfactory performance, explaining 69 to 81% of the variance in fire occurrence. We found that deforestation was consistently the dominant driver of fire across 48 to 70% of the country throughout the study period. Although the influence of low precipitation on fires has increased in 2019 and 2020, the period is not long enough to establish any significant trends. During the study period, wind speed, elevation, and soil moisture had a slight influence of 6–20% without any clear trend, indicating that deforestation continues to be the main driver of fire. Our study improves the current understanding of the drivers of biomass fires across Cambodia, and the methodological framework developed here (quantitative decoupling of the drivers) has strong potential to be applied to other fire-prone areas around the world.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76087656","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Fuqin Li, D. Jupp, B. Markham, I. Lau, C. Ong, G. Byrne, M. Thankappan, Simon Oliver, T. Malthus, P. Fearns
The accuracy of surface reflectance estimation for satellite sensors using radiance-based calibrations can depend significantly on the choice of solar spectral irradiance (or solar spectrum) model used for atmospheric correction. Selecting an accurate solar spectrum model is also important for radiance-based sensor calibration and estimation of atmospheric parameters from irradiance observations. Previous research showed that Landsat 8 could be used to evaluate the quality of solar spectrum models. This paper applies the analysis using five previously evaluated and three more recent solar spectrum models using both Landsat 8 (OLI) and Landsat 9 (OLI2). The study was further extended down to 10 nm resolution and a wavelength range from Ultraviolet A (UVA) to shortwave infrared (SWIR) (370–2480 nm) using inversion of field irradiance measurements. The results using OLI and OLI2 as well as the inversion of irradiance measurements were that the more recent Chance and Kurucz (SA2010), Meftah (SOLAR-ISS) and Coddington (TSIS-1) models performed better than all of the previous models. The results were illustrated by simulating dark and bright surface reflectance signatures obtained by atmospheric correction with the different solar spectrum models. The results showed that if the SA2010 model is assumed to be the “true” solar irradiance, using the TSIS-1 or the SOLAR-ISS model will not significantly change the estimated ground reflectance. The other models differ (some to a large extent) in varying wavelength areas.
{"title":"Choice of Solar Spectral Irradiance Model for Current and Future Remote Sensing Satellite Missions","authors":"Fuqin Li, D. Jupp, B. Markham, I. Lau, C. Ong, G. Byrne, M. Thankappan, Simon Oliver, T. Malthus, P. Fearns","doi":"10.3390/rs15133391","DOIUrl":"https://doi.org/10.3390/rs15133391","url":null,"abstract":"The accuracy of surface reflectance estimation for satellite sensors using radiance-based calibrations can depend significantly on the choice of solar spectral irradiance (or solar spectrum) model used for atmospheric correction. Selecting an accurate solar spectrum model is also important for radiance-based sensor calibration and estimation of atmospheric parameters from irradiance observations. Previous research showed that Landsat 8 could be used to evaluate the quality of solar spectrum models. This paper applies the analysis using five previously evaluated and three more recent solar spectrum models using both Landsat 8 (OLI) and Landsat 9 (OLI2). The study was further extended down to 10 nm resolution and a wavelength range from Ultraviolet A (UVA) to shortwave infrared (SWIR) (370–2480 nm) using inversion of field irradiance measurements. The results using OLI and OLI2 as well as the inversion of irradiance measurements were that the more recent Chance and Kurucz (SA2010), Meftah (SOLAR-ISS) and Coddington (TSIS-1) models performed better than all of the previous models. The results were illustrated by simulating dark and bright surface reflectance signatures obtained by atmospheric correction with the different solar spectrum models. The results showed that if the SA2010 model is assumed to be the “true” solar irradiance, using the TSIS-1 or the SOLAR-ISS model will not significantly change the estimated ground reflectance. The other models differ (some to a large extent) in varying wavelength areas.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79372605","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Y. Cai, Bi-jun Li, Jian Zhou, Hongjuan Zhang, Yongxing Cao
Removing moving objects from 3D LiDAR data plays a crucial role in advancing real-time odometry, life-long SLAM, and motion planning for robust autonomous navigation. In this paper, we present a novel method aimed at addressing the challenges faced by existing approaches when dealing with scenarios involving significant registration errors. The proposed approach offers a unique solution for removing moving objects without the need for registration, leveraging range flow estimation combined with IMU measurements. To this end, our method performs global range flow estimation by utilizing geometric constraints based on the spatio-temporal gradient information derived from the range image, and we introduce IMU measurements to further enhance the accuracy of range flow estimation. Through extensive quantitative evaluations, our approach showcases an improved performance, with an average mIoU of 45.8%, surpassing baseline methods such as Removert (43.2%) and Peopleremover (32.2%). Specifically, it exhibits a substantial improvement in scenarios characterized by a deterioration in registration performance. Moreover, our method does not rely on costly annotations, which make it suitable for SLAM systems with different sensor setups.
{"title":"Removing Moving Objects without Registration from 3D LiDAR Data Using Range Flow Coupled with IMU Measurements","authors":"Y. Cai, Bi-jun Li, Jian Zhou, Hongjuan Zhang, Yongxing Cao","doi":"10.3390/rs15133390","DOIUrl":"https://doi.org/10.3390/rs15133390","url":null,"abstract":"Removing moving objects from 3D LiDAR data plays a crucial role in advancing real-time odometry, life-long SLAM, and motion planning for robust autonomous navigation. In this paper, we present a novel method aimed at addressing the challenges faced by existing approaches when dealing with scenarios involving significant registration errors. The proposed approach offers a unique solution for removing moving objects without the need for registration, leveraging range flow estimation combined with IMU measurements. To this end, our method performs global range flow estimation by utilizing geometric constraints based on the spatio-temporal gradient information derived from the range image, and we introduce IMU measurements to further enhance the accuracy of range flow estimation. Through extensive quantitative evaluations, our approach showcases an improved performance, with an average mIoU of 45.8%, surpassing baseline methods such as Removert (43.2%) and Peopleremover (32.2%). Specifically, it exhibits a substantial improvement in scenarios characterized by a deterioration in registration performance. Moreover, our method does not rely on costly annotations, which make it suitable for SLAM systems with different sensor setups.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76403629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
L. Sfîcă, Alexandru Corocăescu, Claudiu Crețu, Vlad-Alexandru Amihaesei, Pavel Ichim
Using MODIS and Landsat LST images, the present paper advances a series of results on the characteristics of the surface heat island (SUHI) of Bacău City (Romania) during the warm season (April to September) for a period of 20 years (2001–2020). At the same time, given their higher temporal resolution and their availability for both day and night, MODIS LST was used to understand the spatial features of the SUHI in relation to land use. In this way, a total of 946 MODIS Terra and 483 Landsat satellite images were used to outline the main LST characteristics of the days with clear sky in this middle-sized city in northeast Romania. In order to analyze MODIS LST changes in relation to land use changes in the period 2001–2018, we used the standardized CORINE Land Cover datasets. With the help of the Rodionov test, we were able to determine the geometry and intensity of the SUHI. During the day, the spatial extension of the SUHI reaches its maximum level and is delimited by the isotherm of 31.0 °C, which is 1.5–2.0 °C warmer than the neighboring non-urban areas. During the night, the SUHI has a more regulated spatial extension around the central area of the city, delimited by the 15.5 °C isotherm with LST values that are 1.0–1.5 °C warmer than the surrounding non-urban areas. Additionally, from a methodological point of view, we highlight that resampled MODIS and Landsat images at a spatial resolution of 500 m can be used with confidence to understand the detailed spatial features of the SUHI. The results of this study could help the elaboration of future policies meant to mitigate the effects of urbanization on the SUHI in an era of increasing air temperatures during summer.
{"title":"Spatiotemporal Features of the Surface Urban Heat Island of Bacău City (Romania) during the Warm Season and Local Trends of LST Imposed by Land Use Changes during the Last 20 Years","authors":"L. Sfîcă, Alexandru Corocăescu, Claudiu Crețu, Vlad-Alexandru Amihaesei, Pavel Ichim","doi":"10.3390/rs15133385","DOIUrl":"https://doi.org/10.3390/rs15133385","url":null,"abstract":"Using MODIS and Landsat LST images, the present paper advances a series of results on the characteristics of the surface heat island (SUHI) of Bacău City (Romania) during the warm season (April to September) for a period of 20 years (2001–2020). At the same time, given their higher temporal resolution and their availability for both day and night, MODIS LST was used to understand the spatial features of the SUHI in relation to land use. In this way, a total of 946 MODIS Terra and 483 Landsat satellite images were used to outline the main LST characteristics of the days with clear sky in this middle-sized city in northeast Romania. In order to analyze MODIS LST changes in relation to land use changes in the period 2001–2018, we used the standardized CORINE Land Cover datasets. With the help of the Rodionov test, we were able to determine the geometry and intensity of the SUHI. During the day, the spatial extension of the SUHI reaches its maximum level and is delimited by the isotherm of 31.0 °C, which is 1.5–2.0 °C warmer than the neighboring non-urban areas. During the night, the SUHI has a more regulated spatial extension around the central area of the city, delimited by the 15.5 °C isotherm with LST values that are 1.0–1.5 °C warmer than the surrounding non-urban areas. Additionally, from a methodological point of view, we highlight that resampled MODIS and Landsat images at a spatial resolution of 500 m can be used with confidence to understand the detailed spatial features of the SUHI. The results of this study could help the elaboration of future policies meant to mitigate the effects of urbanization on the SUHI in an era of increasing air temperatures during summer.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74633478","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jian Wu, Xiaomei Tang, Long Huang, Shaojie Ni, Feixue Wang
The adaptive beamforming algorithm can realize interference suppression and navigation signal enhancement, and has been widely used. However, achieving high-precision real-time estimation of the direction of arrival (DOA) parameters of navigation signals in strong-interference scenarios with low complexity is still a challenge. In this paper, a blind adaptive beamforming algorithm for a Global Navigation Satellite System (GNSS) array receiver based on direction lock loop is proposed without using the prior information of the DOA parameter. The direction lock loop is used for DOA tracking and estimation after interference suppression, which uses the spatial correlation of the array beam pattern to construct a closed direction-tracking loop. The DOA estimation value is adjusted in real time based on the loop errors. A blind beamformer is constructed using the DOA estimation results to provide gain by forming a beam in the satellite direction. This method improves the accuracy and dynamic adaptability of DOA estimation while significantly reducing the computational complexity. The theoretical analysis and simulation results verify the effectiveness of the proposed algorithm.
{"title":"Blind Adaptive Beamforming for a Global Navigation Satellite System Array Receiver Based on Direction Lock Loop","authors":"Jian Wu, Xiaomei Tang, Long Huang, Shaojie Ni, Feixue Wang","doi":"10.3390/rs15133387","DOIUrl":"https://doi.org/10.3390/rs15133387","url":null,"abstract":"The adaptive beamforming algorithm can realize interference suppression and navigation signal enhancement, and has been widely used. However, achieving high-precision real-time estimation of the direction of arrival (DOA) parameters of navigation signals in strong-interference scenarios with low complexity is still a challenge. In this paper, a blind adaptive beamforming algorithm for a Global Navigation Satellite System (GNSS) array receiver based on direction lock loop is proposed without using the prior information of the DOA parameter. The direction lock loop is used for DOA tracking and estimation after interference suppression, which uses the spatial correlation of the array beam pattern to construct a closed direction-tracking loop. The DOA estimation value is adjusted in real time based on the loop errors. A blind beamformer is constructed using the DOA estimation results to provide gain by forming a beam in the satellite direction. This method improves the accuracy and dynamic adaptability of DOA estimation while significantly reducing the computational complexity. The theoretical analysis and simulation results verify the effectiveness of the proposed algorithm.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78446724","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Large-scale crop mapping is of fundamental importance to tackle food security problems. SAR remote sensing has lately received great attention for crop type mapping due to its stability in the revisit cycle and is not hindered by cloud cover. However, most SAR image-classification studies focused on the application of backscattering characteristics with machine learning models, while few investigated the potential of the polarization decomposition and deep-learning models. This study investigated whether the radar polarization information mined by polarization decomposition, the patch strategy and the approaches for combining recurrent and convolutional neural networks (Conv2d + LSTM and ConvLSTM2d) could effectively improve the accuracy of crop type mapping. Sentinel-1 SLC and GRD products in 2020 were collected as data sources to extract VH, VV, VH/VV, VV + VH, Entropy, Anisotropy, and Alpha 7-dimensional features for classification. The results showed that the three-dimensional Convolutional Neural Network (Conv3d) was the best classifier with an accuracy and kappa up to 88.9% and 0.875, respectively, and the ConvLSTM2d and Conv2d + LSTM achieved the second and third position. Compared to backscatter coefficients, the polarization decomposition features could provide additional phase information for classification in the time dimension. The optimal patch size was 17, and the patch-based Conv3d outperformed the pixel-based Conv1d by 11.3% in accuracy and 0.128 in kappa. This study demonstrated the value of applying polarization decomposition features to deep-learning models and provided a strong technical support to efficient large-scale crop mapping.
{"title":"Crop Type Mapping Based on Polarization Information of Time Series Sentinel-1 Images Using Patch-Based Neural Network","authors":"Yuying Liu, Xuecong Pu, Zhangquan Shen","doi":"10.3390/rs15133384","DOIUrl":"https://doi.org/10.3390/rs15133384","url":null,"abstract":"Large-scale crop mapping is of fundamental importance to tackle food security problems. SAR remote sensing has lately received great attention for crop type mapping due to its stability in the revisit cycle and is not hindered by cloud cover. However, most SAR image-classification studies focused on the application of backscattering characteristics with machine learning models, while few investigated the potential of the polarization decomposition and deep-learning models. This study investigated whether the radar polarization information mined by polarization decomposition, the patch strategy and the approaches for combining recurrent and convolutional neural networks (Conv2d + LSTM and ConvLSTM2d) could effectively improve the accuracy of crop type mapping. Sentinel-1 SLC and GRD products in 2020 were collected as data sources to extract VH, VV, VH/VV, VV + VH, Entropy, Anisotropy, and Alpha 7-dimensional features for classification. The results showed that the three-dimensional Convolutional Neural Network (Conv3d) was the best classifier with an accuracy and kappa up to 88.9% and 0.875, respectively, and the ConvLSTM2d and Conv2d + LSTM achieved the second and third position. Compared to backscatter coefficients, the polarization decomposition features could provide additional phase information for classification in the time dimension. The optimal patch size was 17, and the patch-based Conv3d outperformed the pixel-based Conv1d by 11.3% in accuracy and 0.128 in kappa. This study demonstrated the value of applying polarization decomposition features to deep-learning models and provided a strong technical support to efficient large-scale crop mapping.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74780644","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yu Zhang, Lifu Zhang, Changping Huang, Y. Cen, Q. Tong
The bidirectional reflectance distribution function (BRDF) factor ƒ′ provides a bridge between the inherent and apparent optical properties (IOPs and AOPs) of inland waters. The previous BRDF studies focused on ocean waters, while few studies examine inland waters. It is meaningful to improve the theory of remote sensing of water surface and the accuracy of image derivation in inland waters. In this study, radiative transfer simulation was applied to calculate the ƒ′ values using appropriate IOPs based on in situ combined with realistic boundary conditions (N = 11,232). This study shows that ƒ′ factor varied over the range of 0.33–16.64 in Lake Nansihu, a finite depth water, higher than the range observed for the ocean (0.3–0.6). Our results demonstrate that the factor ƒ′ depends on not only solar zenith angle (θs) but also the average number of collisions (n−) and particulate backscattering ratio (b~bp). The ƒ′ factor shows a continuous geometric increase as the solar zenith angle increases at 400–650 nm but is relatively insensitive to solar angle in the 650–750 nm range in which ƒ′ increases as b~bp and n− decreases. To account for these findings, two empirical models for ƒ′ factor as a function of θs, n− and b~bp are proposed in various spectral wavelengths for Lake Nansihu waters. Our results are crucial for obtaining Hyperspectral normalized reflectance or normalized water-leaving radiance and improving the accuracy of satellite products.
{"title":"Dependence of the Bidirectional Reflectance Distribution Function Factor ƒ′ on the Particulate Backscattering Ratio in an Inland Lake","authors":"Yu Zhang, Lifu Zhang, Changping Huang, Y. Cen, Q. Tong","doi":"10.3390/rs15133392","DOIUrl":"https://doi.org/10.3390/rs15133392","url":null,"abstract":"The bidirectional reflectance distribution function (BRDF) factor ƒ′ provides a bridge between the inherent and apparent optical properties (IOPs and AOPs) of inland waters. The previous BRDF studies focused on ocean waters, while few studies examine inland waters. It is meaningful to improve the theory of remote sensing of water surface and the accuracy of image derivation in inland waters. In this study, radiative transfer simulation was applied to calculate the ƒ′ values using appropriate IOPs based on in situ combined with realistic boundary conditions (N = 11,232). This study shows that ƒ′ factor varied over the range of 0.33–16.64 in Lake Nansihu, a finite depth water, higher than the range observed for the ocean (0.3–0.6). Our results demonstrate that the factor ƒ′ depends on not only solar zenith angle (θs) but also the average number of collisions (n−) and particulate backscattering ratio (b~bp). The ƒ′ factor shows a continuous geometric increase as the solar zenith angle increases at 400–650 nm but is relatively insensitive to solar angle in the 650–750 nm range in which ƒ′ increases as b~bp and n− decreases. To account for these findings, two empirical models for ƒ′ factor as a function of θs, n− and b~bp are proposed in various spectral wavelengths for Lake Nansihu waters. Our results are crucial for obtaining Hyperspectral normalized reflectance or normalized water-leaving radiance and improving the accuracy of satellite products.","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76854338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Francisco Fernandes, Mezgeen A. Rasol, Gilda Schirinzi, Feng Zhou
This Special Issue focuses on the potential of radar-based remote techniques for characterizing and monitoring natural and building structures [...]
本期特刊重点介绍基于雷达的远程技术在表征和监测自然和建筑结构方面的潜力[…]
{"title":"Editorial for the Special Issue \"Radar Techniques for Structures Characterization and Monitoring\"","authors":"Francisco Fernandes, Mezgeen A. Rasol, Gilda Schirinzi, Feng Zhou","doi":"10.3390/rs15133382","DOIUrl":"https://doi.org/10.3390/rs15133382","url":null,"abstract":"This Special Issue focuses on the potential of radar-based remote techniques for characterizing and monitoring natural and building structures [...]","PeriodicalId":20944,"journal":{"name":"Remote. Sens.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76612075","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：