Search

The distribution of industrial heat sources (IHSs) is a crucial indicator for evaluating energy consumption and air pollution levels. However, there is a notable lack of IHS datasets in China that are frequently updated, span long periods, contain detailed characteristic information, have been individually validated and are publicly available. In this study, IHS datasets from China between 2012 and 2021 were constructed using the Visible Infrared Imaging Radiometer Suite (VIIRS) I Band 375 m NRT Active Fire/Hotspots (ACF) Product (VNP14IMGTDL_NRT) to monitor and analyse large‐scale IHSs. First, a density segmentation method based on an improved K‐means algorithm using ACF data and spatial topological correlation analysis was conducted to construct the IHS. Then, 4410 records covering China between 2012 and 2021, with 21 attributes, were obtained and verified, with an individual identification precision of 95.08% via manual verification based on high‐resolution remote‐sensing images and point of interest (POI) data. Finally, the trend of the spatiotemporal variation in IHSs was analysed using a long time series. The results showed that the spatial distribution of IHSs in China from 2012 to 2021 exhibited local aggregation and a gradual shift from east to west. In addition, the number of IHSs in China showed an initial increasing trend from 2012 to 2014, followed by a decrease since 2014, consistent with national energy reform‐related policies. The results of this study indicate the temporal variation in IHSs, enhance the precision of identifying fire location categories and demonstrate the potential for improving energy efficiency, reducing emissions and ensuring sustainable development in China. [ABSTRACT FROM AUTHOR]

The spatiotemporal distribution of industrial heat sources (IHS) is an important indicator for assessing levels of energy consumption and air pollution. Continuous, comprehensive, dynamic monitoring and publicly available datasets of global IHS (GIHS) are lacking and urgently needed. In this study, we built the first long-term (2012–2021) GIHS dataset based on the density-based spatiotemporal clustering method using multi-sources remote sensing data. A total of 25,544 IHS objects with 19 characteristics are identified and validated individually using high-resolution remote sensing images and point of interest (POI) data. The results show that the user's accuracy of the GIHS dataset ranges from 90.95% to 93.46%, surpassing other global IHS products in terms of accuracy, omission rates, and granularity. This long-term GIHS dataset serves as a valuable resource for understanding global environmental changes and making informed policy decisions. Its availability contributes to filling the gap in GIHS data and enhances our knowledge of global-scale industrial heat sources. [ABSTRACT FROM AUTHOR]

The cement industry, as one of the primary contributors to global greenhouse gas emissions, accounts for 7% of the world's carbon dioxide emissions. There is an urgent need to establish a rapid method for detecting cement plants to facilitate effective monitoring. In this study, a comprehensive method based on YOLOv5-IEG and the Thermal Signature Detection module using Google Earth optical imagery and SDGSAT-1 thermal infrared imagery was proposed to detect large-scale cement plant information, including geographic location and operational status. The improved algorithm demonstrated an increase of 4.8% in accuracy and a 7.7% improvement in MAP@.5:95. In a specific empirical investigation in China, we successfully detected 781 large-scale cement plants with an accuracy of 90.8%. Specifically, of the 55 cement plants in Shandong Province, we identified 46 as operational and nine as non-operational. The successful application of advanced models and remote sensing technology in efficiently and accurately tracking the operational status of cement plants provides crucial support for environmental protection and sustainable development. [ABSTRACT FROM AUTHOR]

The significant heterogeneity in the spatial distribution of point of interest (POI) data, the absence of human socio-economic activity information in remote sensing images (RSI), and the high cost of land use (LU) data acquisition restrict their application in PLES spatial identification. Utilizing easily accessible data for detailed spatial identification of PLES remains an urgent challenge, especially when selecting a study area that encompasses both urban built-up areas (UBUA) and non-urban built-up areas (NUBUA). To address this issue, we proposed a PLES spatial identification method that combines POI data and land cover (LC) data in this paper. The proposed method first classified spatial analysis units (SAUs) into agricultural production space (APS), ecological space (ES), and ambiguous space (AS) based on the rich surface physical information from LC data. Subsequently, the AS was further classified into living space (LS) and non-agricultural production space (NAPS) based on the rich human socioeconomic information from POI data. For the AS that contains no POI, a simple rule was established to differentiate it into LS or NAPS. The effectiveness of the method was verified by accuracy evaluation and visual comparison. Applying the method to the Jiaodong Peninsula, we identified the PLES of the Jiaodong Peninsula for 2018 and 2022, further explored their spatial distribution characteristics, and analyzed their changes. Finally, we conducted a discussion on the real-world situations and driving mechanisms of the PLES changes and proposed several policy insights. The results indicated that both the spatial distribution characteristics of PLES and PLES change in the Jiaodong Peninsula were obvious and showed significant differentiation between UBUA and NUBUA. Climatic and natural resource conditions, geographic location, macro-policies, and governmental behaviors drove the PLES changes. [ABSTRACT FROM AUTHOR]

It is of great significance for the sustainable development of steel cities to explore the relationship between the spatial pattern change in steel plants and land cover change during the transformation of steel cities. To address the issue of unsatisfactory results for segmenting steel plants based on high-resolution remote sensing images, due to insufficient sample datasets and task complexity, we proposed a steel plant segmentation strategy that combines high-resolution remote sensing images, POI data, and OSM data. Additionally, we discussed the effect of POI data and OSM data on steel plant segmentation, analyzing the spatial pattern change in steel plants in Tangshan City during 2017–2022 and its relationship with land cover change. The results demonstrate that: (1) The proposed strategy can significantly improve the accuracy of steel plant segmentation. The introduction of POI data can significantly improve the precision of steel plant segmentation, however, it will to some extent reduce the recall of steel plant segmentation, and this phenomenon weakens as the distance threshold increases. The introduction of OSM data can effectively improve the effectiveness of steel plant segmentation, however, it has significant limitations. (2) During 2017–2022, the spatial distribution center of steel plants in Tangshan City moved obviously to the southeast, and the positive change in steel plants was mainly concentrated in the coastal regions of southern Tangshan City, while the negative change in steel plants was mainly concentrated in central Tangshan City. (3) There is a relatively strong spatial correlation between the positive change in steel plants and the transition from vegetation to built area, as well as the transition from cropland to built area. [ABSTRACT FROM AUTHOR]

Considering the classification of high spatial resolution remote sensing imagery, this paper presents a novel classification method for such imagery using deep neural networks. Deep learning methods, such as a fully convolutional network (FCN) model, achieve state-of-the-art performance in natural image semantic segmentation when provided with large-scale datasets and respective labels. To use data efficiently in the training stage, we first pre-segment training images and their labels into small patches as supplements of training data using graph-based segmentation and the selective search method. Subsequently, FCN with atrous convolution is used to perform pixel-wise classification. In the testing stage, post-processing with fully connected conditional random fields (CRFs) is used to refine results. Extensive experiments based on the Vaihingen dataset demonstrate that our method performs better than the reference state-of-the-art networks when applied to high-resolution remote sensing imagery classification. [ABSTRACT FROM AUTHOR]

With the rapid development of satellite remote sensing technology and an ever-increasing number of Earth observation satellites being launched, the global volume of remotely sensed imagery has been growing exponentially. Processing the variety of remotely sensed data has increasingly been complex and difficult. It is also hard to efficiently and intelligently retrieve what users need from a massive database of images. This paper introduces an improved support vector machine (SVM) model, which optimizes the model parameters and selects the feature subset based on the particle swarm optimization (PSO) method and genetic algorithm (GA) for remote sensing image retrieval. The results from an image retrieval experiment show that our method outperforms traditional methods such as GRID, PSO, and GA in terms of consistency and stability. [ABSTRACT FROM PUBLISHER]

Considering the complexity of landslide hazards, their manual investigation lacks efficiency and is time-consuming, especially in high-altitude plateau areas. Therefore, extracting landslide information using remote sensing technology has great advantages. In this study, comprehensive research was carried out on the landslide features of high-resolution remote sensing images on the Mangkam dataset. Based on the idea of feature-driven classification, the landslide extraction model of a fully convolutional spectral–topographic fusion network (FSTF-Net) based on a deep convolutional neural network of multi-source data fusion is proposed, which takes into account the topographic factor (slope and aspect) and the normalized difference vegetation index (NDVI) as multi-source data input by which to train the model. In this paper, a high-resolution remote sensing image classification method based on a fully convolutional network was used to extract the landslide information, thereby realizing the accurate extraction of the landslide and surrounding ground-object information. With Mangkam County in the southeast of the Qinghai–Tibet Plateau China as the study area, the proposed method was evaluated based on the high-precision digital elevation model (DEM) generated from stereoscopic images of Resources Satellite-3 and multi-source high-resolution remote sensing image data (Beijing-2, Worldview-3, and SuperView-1). Results show that our method had a landslide detection precision of 0.85 and an overall classification accuracy of 0.89. Compared with the latest DeepLab_v3+, our model increases the landslide detection precision by 5%. Thus, the proposed FSTF-Net model has high reliability and robustness. [ABSTRACT FROM AUTHOR]

Class imbalance is a key issue for the application of deep learning for remote sensing image classification because a model generated by imbalanced samples training has low classification accuracy for minority classes. In this study, an accurate classification approach using the multistage sampling method and deep neural networks was proposed to classify imbalanced data. We first balance samples by multistage sampling to obtain the training sets. Then, a state-of-the-art model is adopted by combining the advantages of atrous spatial pyramid pooling (ASPP) and Encoder-Decoder for pixel-wise classification, which are two different types of fully convolutional networks (FCNs) that can obtain contextual information of multiple levels in the Encoder stage. The details and spatial dimensions of targets are restored using such information during the Decoder stage. We employ four deep learning-based classification algorithms (basic FCN, FCN-8S, ASPP, and Encoder-Decoder with ASPP of our approach) on multistage training sets (original, MUS1, and MUS2) of WorldView-3 images in southeastern Qinghai-Tibet Plateau and GF-2 images in northeastern Beijing for comparison. The experiments show that, compared with existing sets (original, MUS1, and identical) and existing method (cost weighting), the MUS2 training set of multistage sampling significantly enhance the classification performance for minority classes. Our approach shows distinct advantages for imbalanced data. [ABSTRACT FROM AUTHOR]