Pub Date : 2023-10-23DOI: 10.1080/2150704x.2023.2270107
Ibnu F. Kurniawan, Fei He, Iswan Dunggio, Marini S. Hamidun, Zulham Sirajuddin, Muhammad Aziz, A. Taufiq Asyhari
ABSTRACTRemote sensing technologies have been increasingly crucial to support policy-makers in achieving their ecological strategies. The data provided by such technology can estimate the bioenergy source production rate and monitor deforestation. This work participates in the cause by contributing an aerial dataset and developing an intelligent tree-detection system usable for counting trees with the bioenergy potential. Low-altitude flying units have been vastly used for such a purpose due to their ability to capture high-quality data from distant locations. Despite these potentials, collected images that compose a dataset are often characterized by imbalanced distribution among classes. The class disproportion can affect the overall model performance, as it severely deprives key features of under-represented classes. This study proposes data-level approaches that adopt and extend prior sampling algorithms for object detection problems. The devised techniques try to reduce the number of redundant outputs obtained from sampling methods and reduce the iteration required to achieve the target imbalance ratio by employing a systematic flow. In such a process, the class distribution of an original dataset is used as a guideline for selecting candidates for subsequent processes. Our results show that the modified dataset can reduce the length of a training process shown by fewer iterations required to achieve the final metrics of its original dataset version and lower training losses in each iteration. Additionally, the modified dataset can improve the F-score (F1) and precision metric of object detection algorithm by up to 6%.KEYWORDS: Aerial surveillanceUrban forestryRemote monitoringClass imbalancedObject detectionMachine learning Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported in part by the British Council COP26 Trilateral Research Initiative grant under the project ”Scaling-up Indonesian Bioenergy Potential through Assessment of Wallacea’s Plant Species: Data-Driven Energy Harvesting and Community-Centred Approach”. Ibnu F. Kurniawan acknowledged the support from the Directorate General of Higher Education, Research, and Technology, Indonesia.
遥感技术在支持决策者实现其生态战略方面发挥着越来越重要的作用。这种技术提供的数据可以估计生物能源的生产速度和监测森林砍伐。这项工作通过提供航空数据集和开发可用于计算具有生物能源潜力的树木的智能树木检测系统来参与这项事业。由于低空飞行单位能够从遥远地点捕获高质量数据，因此已广泛用于这一目的。尽管有这些潜力，收集到的图像组成的数据集往往具有类之间分布不平衡的特点。类比例失调会影响整体模型性能，因为它严重剥夺了代表性不足的类的关键特征。本研究提出了数据级方法，采用并扩展了目标检测问题的先验采样算法。所设计的技术试图减少从采样方法中获得的冗余输出的数量，并通过采用系统流程减少实现目标不平衡比所需的迭代。在这个过程中，原始数据集的类分布被用作后续过程选择候选的指导方针。我们的研究结果表明，修改后的数据集可以减少训练过程的长度，通过更少的迭代来达到原始数据集版本的最终指标，并且减少每次迭代的训练损失。此外，改进后的数据集可将目标检测算法的F-score (F1)和精度指标提高6%。关键词:航空监测城市林业远程监测类失衡对象检测机器学习披露声明作者未报告潜在的利益冲突。这项工作得到了英国文化协会COP26三边研究倡议项目“通过评估Wallacea植物物种扩大印度尼西亚生物能源潜力:数据驱动的能源收集和以社区为中心的方法”的部分资助。Ibnu F. Kurniawan感谢印度尼西亚高等教育、研究和技术总局的支持。
{"title":"Imbalanced learning of remotely sensed data for bioenergy source identification in a forest in the Wallacea region of Indonesia","authors":"Ibnu F. Kurniawan, Fei He, Iswan Dunggio, Marini S. Hamidun, Zulham Sirajuddin, Muhammad Aziz, A. Taufiq Asyhari","doi":"10.1080/2150704x.2023.2270107","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2270107","url":null,"abstract":"ABSTRACTRemote sensing technologies have been increasingly crucial to support policy-makers in achieving their ecological strategies. The data provided by such technology can estimate the bioenergy source production rate and monitor deforestation. This work participates in the cause by contributing an aerial dataset and developing an intelligent tree-detection system usable for counting trees with the bioenergy potential. Low-altitude flying units have been vastly used for such a purpose due to their ability to capture high-quality data from distant locations. Despite these potentials, collected images that compose a dataset are often characterized by imbalanced distribution among classes. The class disproportion can affect the overall model performance, as it severely deprives key features of under-represented classes. This study proposes data-level approaches that adopt and extend prior sampling algorithms for object detection problems. The devised techniques try to reduce the number of redundant outputs obtained from sampling methods and reduce the iteration required to achieve the target imbalance ratio by employing a systematic flow. In such a process, the class distribution of an original dataset is used as a guideline for selecting candidates for subsequent processes. Our results show that the modified dataset can reduce the length of a training process shown by fewer iterations required to achieve the final metrics of its original dataset version and lower training losses in each iteration. Additionally, the modified dataset can improve the F-score (F1) and precision metric of object detection algorithm by up to 6%.KEYWORDS: Aerial surveillanceUrban forestryRemote monitoringClass imbalancedObject detectionMachine learning Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported in part by the British Council COP26 Trilateral Research Initiative grant under the project ”Scaling-up Indonesian Bioenergy Potential through Assessment of Wallacea’s Plant Species: Data-Driven Energy Harvesting and Community-Centred Approach”. Ibnu F. Kurniawan acknowledged the support from the Directorate General of Higher Education, Research, and Technology, Indonesia.","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135412695","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-10-03DOI: 10.1080/2150704x.2023.2266118
Aiye Shi, Ziqi Li, Xin Wang
ABSTRACTRemote sensing image (RSI) scene classification is a hot topic in the field of remote sensing and has garnered a lot of attention. The key issue in image classification is effectively understanding semantic content. Convolutional neural networks (CNNs) are generally recognized to significantly improve classification performance due to their powerful feature extraction capabilities. However, the overall structure of the model is complicated and has a large number of parameters, making it difficult to extract more efficient features. To address these problems, in this paper, we propose a lightweight skip-connected expansion Inception network called SEINet. To capture characteristics at a more granular level, we create a new lightweight backbone network with fewer parameters based on the existing network architecture. Additionally, the paper introduces a skip-connected expansion Inception (SEI) module for extracting context-dependent relationships. The ablation experiments verify the effectiveness of our proposed module. Experiment findings on two public datasets demonstrate that our method has advantages in classification accuracy and execution efficiency over state-of-the-art (SOTA) methods.KEYWORDS: Remote sensingscene classificationconvolution neural network (CNN)skip-connected expansion Inception Disclosure statementNo potential conflict of interest was reported by the author(s).
{"title":"A lightweight skip-connected expansion inception network for remote sensing scene classification","authors":"Aiye Shi, Ziqi Li, Xin Wang","doi":"10.1080/2150704x.2023.2266118","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2266118","url":null,"abstract":"ABSTRACTRemote sensing image (RSI) scene classification is a hot topic in the field of remote sensing and has garnered a lot of attention. The key issue in image classification is effectively understanding semantic content. Convolutional neural networks (CNNs) are generally recognized to significantly improve classification performance due to their powerful feature extraction capabilities. However, the overall structure of the model is complicated and has a large number of parameters, making it difficult to extract more efficient features. To address these problems, in this paper, we propose a lightweight skip-connected expansion Inception network called SEINet. To capture characteristics at a more granular level, we create a new lightweight backbone network with fewer parameters based on the existing network architecture. Additionally, the paper introduces a skip-connected expansion Inception (SEI) module for extracting context-dependent relationships. The ablation experiments verify the effectiveness of our proposed module. Experiment findings on two public datasets demonstrate that our method has advantages in classification accuracy and execution efficiency over state-of-the-art (SOTA) methods.KEYWORDS: Remote sensingscene classificationconvolution neural network (CNN)skip-connected expansion Inception Disclosure statementNo potential conflict of interest was reported by the author(s).","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135695514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-10-03DOI: 10.1080/2150704x.2023.2266119
Dong Li, Hengbiao Zheng, Xia Yao, Yan Zhu, Weixing Cao, Tao Cheng
ABSTRACTThe canopy chlorophyll content (CCC) provides valuable information about the crop growth status. CCC can be estimated using remote sensing techniques, such as through the red-edge-based chlorophyll index (CIRE). The empirical model between CCC and CIRE calibrated using the measured dataset lacks generality. Therefore, the semi-empirical model is a better choice, which is calibrated on the physical model simulations. However, the effect of parameter settings of physical models on semi-empirical models is not clear. This study first investigated the effects of dry matter content (LMA) and mesophyll structural coefficient (Ns) on the CCC-CIRE relationships and then evaluated CCC estimation using the CIRE-based semi-empirical model calibrated on simulated datasets with different ranges of LMA and Ns. The results showed that the relationships between CCC and CIRE were sensitive to Ns and LMA. Therefore, after considering the prior information of Ns (1.0–1.5) and LMA (20–80 g m−2) for the crop, the best estimation of CCC was obtained with an R2 of 0.82 and an RMSE of 0.36 g m−2, which were substantially better than the model without considering the prior information (R2 = 0.40 and RMSE = 0.67 g m−2). These findings improved our understanding of CCC estimation using the semi-empirical model and would facilitate the accurate mapping of CCC for agricultural management.KEYWORDS: canopy chlorophyll contentvegetation indexsemi-empirical modelprior information Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementThe data used in this study are available upon request.Additional informationFundingThis work was supported by grants from the National Natural Science Foundation of China (42101360, 32021004), the Jiangsu Funding Program for Excellent Postdoctoral Talent (2022ZB333), the Fellowship of China Postdoctoral Science Foundation (2022M710070), and Collaborative Innovation Center for Modern Crop Production co-sponsored by Province and Ministry. We are grateful to the reviewers for their suggestions and comments, which significantly improved the quality of this paper.
摘要冠层叶绿素含量(CCC)是反映作物生长状况的重要信息。可以利用遥感技术，如基于红边的叶绿素指数(CIRE)来估算CCC。使用实测数据校准的CCC与CIRE之间的经验模型缺乏通用性。因此，在物理模型模拟的基础上进行标定的半经验模型是较好的选择。然而，物理模型参数设置对半经验模型的影响尚不清楚。本研究首先探讨了干物质含量(LMA)和叶肉结构系数(Ns)对CCC- cire关系的影响，然后利用基于cre的半经验模型在不同LMA和Ns范围的模拟数据集上进行了校准。结果表明，CCC和CIRE之间的关系对Ns和LMA敏感。因此，在考虑作物的Ns(1.0-1.5)和LMA (20-80 g m−2)的先验信息后，获得了最佳的CCC估计，R2为0.82,RMSE为0.36 g m−2，大大优于不考虑先验信息的模型(R2 = 0.40, RMSE = 0.67 g m−2)。这些发现提高了我们对使用半经验模型估算CCC的理解，并将有助于农业管理中CCC的准确定位。关键词:冠层叶绿素含量植被指数半经验模型先验信息披露声明作者未报告潜在利益冲突。数据可用性声明本研究中使用的数据可应要求提供。项目资助:国家自然科学基金项目(42101360,32021004)、江苏省优秀博士后人才资助项目(2022ZB333)、中国博士后科学基金项目(2022M710070)和省部级现代作物生产协同创新中心。感谢审稿人提出的建议和意见，大大提高了本文的质量。
{"title":"Semi-empirical models for estimating canopy chlorophyll content: the importance of prior information","authors":"Dong Li, Hengbiao Zheng, Xia Yao, Yan Zhu, Weixing Cao, Tao Cheng","doi":"10.1080/2150704x.2023.2266119","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2266119","url":null,"abstract":"ABSTRACTThe canopy chlorophyll content (CCC) provides valuable information about the crop growth status. CCC can be estimated using remote sensing techniques, such as through the red-edge-based chlorophyll index (CIRE). The empirical model between CCC and CIRE calibrated using the measured dataset lacks generality. Therefore, the semi-empirical model is a better choice, which is calibrated on the physical model simulations. However, the effect of parameter settings of physical models on semi-empirical models is not clear. This study first investigated the effects of dry matter content (LMA) and mesophyll structural coefficient (Ns) on the CCC-CIRE relationships and then evaluated CCC estimation using the CIRE-based semi-empirical model calibrated on simulated datasets with different ranges of LMA and Ns. The results showed that the relationships between CCC and CIRE were sensitive to Ns and LMA. Therefore, after considering the prior information of Ns (1.0–1.5) and LMA (20–80 g m−2) for the crop, the best estimation of CCC was obtained with an R2 of 0.82 and an RMSE of 0.36 g m−2, which were substantially better than the model without considering the prior information (R2 = 0.40 and RMSE = 0.67 g m−2). These findings improved our understanding of CCC estimation using the semi-empirical model and would facilitate the accurate mapping of CCC for agricultural management.KEYWORDS: canopy chlorophyll contentvegetation indexsemi-empirical modelprior information Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementThe data used in this study are available upon request.Additional informationFundingThis work was supported by grants from the National Natural Science Foundation of China (42101360, 32021004), the Jiangsu Funding Program for Excellent Postdoctoral Talent (2022ZB333), the Fellowship of China Postdoctoral Science Foundation (2022M710070), and Collaborative Innovation Center for Modern Crop Production co-sponsored by Province and Ministry. We are grateful to the reviewers for their suggestions and comments, which significantly improved the quality of this paper.","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135739363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-30DOI: 10.1080/2150704x.2023.2264491
Mohamed Baadeche, Faouzi Soltani
ABSTRACTIn a radar detection system, multiple pulse (MP) transmission is used to improve detection performance compared to the single pulse case by integrating the echoes of pulses at reception. In this paper, we derive closed-form expressions of the probability of false alarm PFA of the cell averaging-constant false alarm rate, greatest-of CFAR, and smallest-of CFAR detectors considering a homogeneous gamma distributed radar clutter applied to the MP case. Expressions are given by analytical formulas for a positive real shape parameter which correspond to a real situation and are validated by comparing them in terms of the detection threshold calculated values T, to the results obtained by means of Monte Carlo simulations.KEYWORDS: Multiple pulsesCA-CFARGO-CFARSO-CFARgamma distributed clutter Disclosure statementNo potential conflict of interest was reported by the author(s).
{"title":"Closed-form expressions of <i>P</i> <sub>FA</sub> of mean level CFAR detectors for multiple-pulse gamma-distributed radar clutter","authors":"Mohamed Baadeche, Faouzi Soltani","doi":"10.1080/2150704x.2023.2264491","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2264491","url":null,"abstract":"ABSTRACTIn a radar detection system, multiple pulse (MP) transmission is used to improve detection performance compared to the single pulse case by integrating the echoes of pulses at reception. In this paper, we derive closed-form expressions of the probability of false alarm PFA of the cell averaging-constant false alarm rate, greatest-of CFAR, and smallest-of CFAR detectors considering a homogeneous gamma distributed radar clutter applied to the MP case. Expressions are given by analytical formulas for a positive real shape parameter which correspond to a real situation and are validated by comparing them in terms of the detection threshold calculated values T, to the results obtained by means of Monte Carlo simulations.KEYWORDS: Multiple pulsesCA-CFARGO-CFARSO-CFARgamma distributed clutter Disclosure statementNo potential conflict of interest was reported by the author(s).","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136279775","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-30DOI: 10.1080/2150704x.2023.2264494
N Bidyarani Chanu, Bakimchandra Oinam
ABSTRACTRice is a staple food for the vast majority of the world’s population and one of the world’s largest consumers of freshwater. Unfortunately, climate change will further worsen the demand for blue water demand, particularly for rice cultivation needs to be closely monitored. Our study assessed the spatial water footprint (WF) of rice for the valley region of Manipur using Moderate Resolution Imaging Spectro-radiometer Evapotranspiration (MOD16) and Climate Hazards Group Infrared Precipitation with Station (CHIRPS) datasets. In addition, rice’s economic water productivity of green and blue water was evaluated. Results showed an average spatial WF ranging from 772.14 to 1456.23 m3/tonne. According to data comparing the national average, 90% of the valley area has a lower WF value than the country as a whole. The green and blue WF of rice ranges from 596.62 m3/tonne to 673.42 m3/tonne and 65.79 m3/tonne to 767.65 m3/tonne, respectively. The spatial variation of the blue WF is due to the amount of rainfall and irrigation application within the study area. The green economic water productivity is getting lower than the blue economic water productivity due to green water’s lesser economic scarcity than blue water. This study can help plan crop allocation in favour of water availability by water management authorities on economic value calculations.KEYWORDS: MOD16CHIRPSricewater footprint AcknowledgmentsWe thank the Department of Agriculture, Manipur, and the Directorate of Environment and Climate Change, Manipur, for providing crop-related and weather data for running this project. We also thank NASA and CHIRPS for providing the dataset through the respective archives. We will be grateful to the Ministry of Human Resource Development, the Government of India and the National Institute of Technology Manipur for PhD fellowship.Disclosure statementNo potential conflict of interest was reported by the authors.
{"title":"The impact of rice cultivation in green and blue water on the economic productivity of the valley region of Manipur, India","authors":"N Bidyarani Chanu, Bakimchandra Oinam","doi":"10.1080/2150704x.2023.2264494","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2264494","url":null,"abstract":"ABSTRACTRice is a staple food for the vast majority of the world’s population and one of the world’s largest consumers of freshwater. Unfortunately, climate change will further worsen the demand for blue water demand, particularly for rice cultivation needs to be closely monitored. Our study assessed the spatial water footprint (WF) of rice for the valley region of Manipur using Moderate Resolution Imaging Spectro-radiometer Evapotranspiration (MOD16) and Climate Hazards Group Infrared Precipitation with Station (CHIRPS) datasets. In addition, rice’s economic water productivity of green and blue water was evaluated. Results showed an average spatial WF ranging from 772.14 to 1456.23 m3/tonne. According to data comparing the national average, 90% of the valley area has a lower WF value than the country as a whole. The green and blue WF of rice ranges from 596.62 m3/tonne to 673.42 m3/tonne and 65.79 m3/tonne to 767.65 m3/tonne, respectively. The spatial variation of the blue WF is due to the amount of rainfall and irrigation application within the study area. The green economic water productivity is getting lower than the blue economic water productivity due to green water’s lesser economic scarcity than blue water. This study can help plan crop allocation in favour of water availability by water management authorities on economic value calculations.KEYWORDS: MOD16CHIRPSricewater footprint AcknowledgmentsWe thank the Department of Agriculture, Manipur, and the Directorate of Environment and Climate Change, Manipur, for providing crop-related and weather data for running this project. We also thank NASA and CHIRPS for providing the dataset through the respective archives. We will be grateful to the Ministry of Human Resource Development, the Government of India and the National Institute of Technology Manipur for PhD fellowship.Disclosure statementNo potential conflict of interest was reported by the authors.","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136341623","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-30DOI: 10.1080/2150704x.2023.2264493
Wenqing Feng, Jihui Tu, Chenhao Sun, Wei Xu
ABSTRACTRemote sensing change detection (CD) methods that rely on supervised deep convolutional neural networks require large-scale labelled data, which is time-consuming and laborious to collect and label, especially for bi-temporal samples containing changed areas. Conversely, acquiring a large volume of unannotated images is relatively easy. Recently, self-supervised contrastive learning has emerged as a promising method for learning from unannotated images, thereby reducing the need for annotation. However, most existing methods employ random values or ImageNet pre-trained models to initialize their encoders and lack prior knowledge tailored to the demands of CD tasks, thus constraining the performance of CD models. To address these challenges, we propose a novel Barlow Twins self-supervised pre-training method for CD (BTSCD), which uses absolute feature differences to directly learn distinct representations associated with changed regions from unlabelled bi-temporal remote sensing images in a self-supervised manner. Experimental results obtained using two publicly available CD datasets demonstrate that our proposed approach exhibits competitive quantitative performance. Moreover, the proposed method achieved final results superior to those of existing state-of-the-art methods. Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported by the National Natural Science Foundation of China under Grant Nos. 42101358.
{"title":"Barlow twin self-supervised pre-training for remote sensing change detection","authors":"Wenqing Feng, Jihui Tu, Chenhao Sun, Wei Xu","doi":"10.1080/2150704x.2023.2264493","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2264493","url":null,"abstract":"ABSTRACTRemote sensing change detection (CD) methods that rely on supervised deep convolutional neural networks require large-scale labelled data, which is time-consuming and laborious to collect and label, especially for bi-temporal samples containing changed areas. Conversely, acquiring a large volume of unannotated images is relatively easy. Recently, self-supervised contrastive learning has emerged as a promising method for learning from unannotated images, thereby reducing the need for annotation. However, most existing methods employ random values or ImageNet pre-trained models to initialize their encoders and lack prior knowledge tailored to the demands of CD tasks, thus constraining the performance of CD models. To address these challenges, we propose a novel Barlow Twins self-supervised pre-training method for CD (BTSCD), which uses absolute feature differences to directly learn distinct representations associated with changed regions from unlabelled bi-temporal remote sensing images in a self-supervised manner. Experimental results obtained using two publicly available CD datasets demonstrate that our proposed approach exhibits competitive quantitative performance. Moreover, the proposed method achieved final results superior to those of existing state-of-the-art methods. Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported by the National Natural Science Foundation of China under Grant Nos. 42101358.","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136341616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-30DOI: 10.1080/2150704x.2023.2264492
Xiao Tang, Chenlu Li, Yongxing Du, Ling Qin, Baoshan Li
ABSTRACTThe carrier platform in synthetic aperture radar (SAR) imaging may deviate from the correct trajectory due to end currents in the atmosphere, causing motion errors and ultimately degrading the quality of the radar image. Techniques for motion compensation can reduce the impact of motion errors on the results of the imaging process. Motion error parameters are necessary for motion compensation algorithms. In this paper, the line of sight (LOS) error of the carrier platform is estimated based on the vortex-selected SAR imaging system by analysing the vortex SAR echo data using the relationship between the Bessel magnitude term, the phase term, and the motion error. The validity of the method is verified through numerical simulations.KEYWORDS: Electromagnetic vortexmotion error estimationSAR imaging Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingNational Natural Science Foundation of China (61961033).
{"title":"Motion error parameter estimation based on vortex echo data","authors":"Xiao Tang, Chenlu Li, Yongxing Du, Ling Qin, Baoshan Li","doi":"10.1080/2150704x.2023.2264492","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2264492","url":null,"abstract":"ABSTRACTThe carrier platform in synthetic aperture radar (SAR) imaging may deviate from the correct trajectory due to end currents in the atmosphere, causing motion errors and ultimately degrading the quality of the radar image. Techniques for motion compensation can reduce the impact of motion errors on the results of the imaging process. Motion error parameters are necessary for motion compensation algorithms. In this paper, the line of sight (LOS) error of the carrier platform is estimated based on the vortex-selected SAR imaging system by analysing the vortex SAR echo data using the relationship between the Bessel magnitude term, the phase term, and the motion error. The validity of the method is verified through numerical simulations.KEYWORDS: Electromagnetic vortexmotion error estimationSAR imaging Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingNational Natural Science Foundation of China (61961033).","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136279602","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
ABSTRACTTotal ozone time series in a metropolis in India were examined in the current study. The data explored are obtained through Brewer spectrophotometer, which counts photons with a photomultiplier to calculate UV irradiance in the spectrum and measures total ozone when the relative route of photons through the ozone layer (air mass) is 3.5 or less. The total ozone time series’ uncertainty was thoroughly examined using the Dempster–Shafer method, and the association was also depicted using three-dimensional graphs. Finally, the Adam Optimisation Algorithm and the Rectified Linear Unit were used to demonstrate the prediction capability of the single layer Long Short-Term Memory model.KEYWORDS: Total ozoneDempster–Shafer theoryfuzzy setjoint belief measureLSTM Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementThe World Ozone and Ultraviolet Radiation Data Centre (WOUDC) website was used to collect the observational Total Ozone monthly data for the region of New Delhi, India, for the years 2017 to 2021. The data can be accessed at the following link: https://woudc.org/data/explore.php#.
{"title":"Dempster–Shafer and LSTM based analysis and forecasting of total ozone data","authors":"Rashmi Rekha Devi, Soumya Banerjee, Surajit Chattopadhyay","doi":"10.1080/2150704x.2023.2258466","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2258466","url":null,"abstract":"ABSTRACTTotal ozone time series in a metropolis in India were examined in the current study. The data explored are obtained through Brewer spectrophotometer, which counts photons with a photomultiplier to calculate UV irradiance in the spectrum and measures total ozone when the relative route of photons through the ozone layer (air mass) is 3.5 or less. The total ozone time series’ uncertainty was thoroughly examined using the Dempster–Shafer method, and the association was also depicted using three-dimensional graphs. Finally, the Adam Optimisation Algorithm and the Rectified Linear Unit were used to demonstrate the prediction capability of the single layer Long Short-Term Memory model.KEYWORDS: Total ozoneDempster–Shafer theoryfuzzy setjoint belief measureLSTM Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementThe World Ozone and Ultraviolet Radiation Data Centre (WOUDC) website was used to collect the observational Total Ozone monthly data for the region of New Delhi, India, for the years 2017 to 2021. The data can be accessed at the following link: https://woudc.org/data/explore.php#.","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135967378","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-21DOI: 10.1080/2150704x.2023.2261150
Zhixuan Zhou, Weining Zhu
ABSTRACTThe spectral probability distributions (SPD) of water bodies in satellite images have demonstrated the potential for indicating the geographical and environmental features of their watersheds. This implies that SPDs also have the potential for indicating water quality features, but so far there have been no further studies on their correlations. In this study, 690 SPDs of global closed connected water bodies, mainly including lakes and reservoirs, were extracted from Landsat-8 images. These SPDs were classified into seven types, and the entropy of each SPD diagram was calculated. The correlation between the SPD diagram’s entropy and Forel-Ule index (FUI) is relatively good with R2 = 0.5651 – indicating that water bodies with better water quality are usually found to have smaller entropy in their SPD diagrams. This study demonstrates that SPD is a good indicator for not only the aquatic environment but also water quality monitoring.KEYWORDS: Forel-Ule index (FUI)spectral probability distribution (SPD)Landsat-8remote sensingwater quality AcknowledgmentsThis research was funded by the National Natural Science Foundation of China (No. 41971373) and the Science Foundation of Donghai Laboratory (No. DH-2022KF01009).Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThe work was supported by the National Natural Science Foundation of China [41971373]; Science Foundation of Donghai Laboratory [DH-2022KF01009].
{"title":"Water quality indication of spectral probability distribution (SPD): correlation between SPD and Forel-Ule index in closed, connected water bodies","authors":"Zhixuan Zhou, Weining Zhu","doi":"10.1080/2150704x.2023.2261150","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2261150","url":null,"abstract":"ABSTRACTThe spectral probability distributions (SPD) of water bodies in satellite images have demonstrated the potential for indicating the geographical and environmental features of their watersheds. This implies that SPDs also have the potential for indicating water quality features, but so far there have been no further studies on their correlations. In this study, 690 SPDs of global closed connected water bodies, mainly including lakes and reservoirs, were extracted from Landsat-8 images. These SPDs were classified into seven types, and the entropy of each SPD diagram was calculated. The correlation between the SPD diagram’s entropy and Forel-Ule index (FUI) is relatively good with R2 = 0.5651 – indicating that water bodies with better water quality are usually found to have smaller entropy in their SPD diagrams. This study demonstrates that SPD is a good indicator for not only the aquatic environment but also water quality monitoring.KEYWORDS: Forel-Ule index (FUI)spectral probability distribution (SPD)Landsat-8remote sensingwater quality AcknowledgmentsThis research was funded by the National Natural Science Foundation of China (No. 41971373) and the Science Foundation of Donghai Laboratory (No. DH-2022KF01009).Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThe work was supported by the National Natural Science Foundation of China [41971373]; Science Foundation of Donghai Laboratory [DH-2022KF01009].","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136130533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-21DOI: 10.1080/2150704x.2023.2258460
Ulf Norinder, Stephanie Lowry
This investigation shows that successful forecasting models for monitoring forest health status with respect to Larch Casebearer damages can be derived using a combination of a confidence predictor framework (Conformal Prediction) in combination with a deep learning architecture (Yolo v5). A confidence predictor framework can predict the current types of diseases used to develop the model and also provide indication of new, unseen, types or degrees of disease. The user of the models is also, at the same time, provided with reliable predictions and a well-established applicability domain for the model where such reliable predictions can and cannot be expected. Furthermore, the framework gracefully handles class imbalances without explicit over- or under-sampling or category weighting which may be of crucial importance in cases of highly imbalanced datasets. The present approach also provides indication of when insufficient information has been provided as input to the model at the level of accuracy (reliability) need by the user to make subsequent decisions based on the model predictions.
{"title":"Predicting Larch Casebearer damage with confidence using Yolo network models and conformal prediction","authors":"Ulf Norinder, Stephanie Lowry","doi":"10.1080/2150704x.2023.2258460","DOIUrl":"https://doi.org/10.1080/2150704x.2023.2258460","url":null,"abstract":"This investigation shows that successful forecasting models for monitoring forest health status with respect to Larch Casebearer damages can be derived using a combination of a confidence predictor framework (Conformal Prediction) in combination with a deep learning architecture (Yolo v5). A confidence predictor framework can predict the current types of diseases used to develop the model and also provide indication of new, unseen, types or degrees of disease. The user of the models is also, at the same time, provided with reliable predictions and a well-established applicability domain for the model where such reliable predictions can and cannot be expected. Furthermore, the framework gracefully handles class imbalances without explicit over- or under-sampling or category weighting which may be of crucial importance in cases of highly imbalanced datasets. The present approach also provides indication of when insufficient information has been provided as input to the model at the level of accuracy (reliability) need by the user to make subsequent decisions based on the model predictions.","PeriodicalId":49132,"journal":{"name":"Remote Sensing Letters","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136235358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","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：