In recent years, remote sensing target recognition algorithms based on deep learning technology have gradually become mainstream in the field of remote sensing because of the great improvements that have been made in the accuracy of image target recognition through the use of deep learning. In the research of remote sensing image target recognition based on deep learning, an insufficient number of research samples is often an encountered issue; too small a number of research samples will cause the phenomenon of an overfitting of the model. To solve this problem, data augmentation techniques have also been developed along with the popularity of deep learning, and many methods have been proposed. However, to date, there is no literature aimed at expounding and summarizing the current state of the research applied to data augmentation for remote sensing object recognition, which is the purpose of this article. First, based on the essential principles of data augmentation methods, the existing methods are divided into two categories: data-based data augmentation methods and network-based data augmentation methods. Second, this paper subdivides and compares each method category to show the advantages, disadvantages, and characteristics of each method. Finally, this paper discusses the limitations of the existing methods and points out future research directions for data augmentation methods.

With the development and improvement of modern surveying and remote-sensing technology, data in the fields of surveying and remote sensing have grown rapidly. Due to the characteristics of large-scale, heterogeneous and diverse surveys and the loose organization of surveying and remote-sensing data, effectively obtaining information and knowledge from data can be difficult. Therefore, this paper proposes a method of using ontology for heterogeneous data integration. Based on the heterogeneous, decentralized, and dynamic updates of large surveying and remote-sensing data, this paper constructs a knowledge graph for surveying and remote-sensing applications. First, data are extracted. Second, using the ontology editing tool Protégé, a knowledge graph mode level is constructed. Then, using a relational database, data are stored, and a D2RQ tool maps the data from the mode level’s ontology to the data layer. Then, using the D2RQ tool, a SPARQL protocol and resource description framework query language (SPARQL) endpoint service is used to describe functions such as query and reasoning of the knowledge graph. The graph database is then used to display the knowledge graph. Finally, the knowledge graph is used to describe the correlation between the fields of surveying and remote sensing.

Sentinel-2 NDVI and surface reflectance time series have been widely used in various geoscience research, but the data is deteriorated or missing due to the cloud contamination, so it is necessary to reconstruct the Sentinel-2 NDVI and surface reflectance time series. At present, there are few studies on reconstructing the Sentinel-2 NDVI or surface reflectance time series, and these existing reconstruction methods have some shortcomings. We proposed a new method to reconstruct the Sentinel-2 NDVI and surface reflectance time series using the penalized least-square regression based on discrete cosine transform (DCT-PLS) method. This method iteratively identifies cloud-contaminated NDVI over NDVI time series from the Sentinel-2 surface reflectance data by adjusting the weights. The NDVI and surface reflectance time series are then reconstructed from cloud-free NDVI and surface reflectance using the adjusted weights as constraints. We have made some improvements to the DCT-PLS method. First, the traditional discrete cosine transformation (DCT) in the DCT-PLS method is matrix generated from discrete and equally spaced data, we reconfigured the DCT formulas to adapt for irregular interval time series, and optimized the control parameters $N$ and $s$ according to the typical vegetation samples in China. Second, the DCT-PLS method was deployed in the Google Earth Engine (GEE) platform for the efficiency and convenience of data users. We used the DCT-PLS method to reconstruct the Sentinel-2 NDVI time series and surface reflectance time series in the blue, green, red, and near infrared (NIR) bands in typical vegetation samples and the Zhangjiakou and Hangzhou study area. We found that this method performed better than the SG filter method in reconstructing the NDVI time series, and can identify and reconstruct the contaminated NDVI as well as surface reflectance with low root mean square error (RMSE) and high coefficient of determination (R²). However, in cases of a long range of cloud contamination, or above water surface, it may be necessary to increase the control parameter $s$ for a more stable performance. The GEE code is freely available online and the link is in the conclusions of this article, researchers are welcome to use this method to generate cloudless Sentinel-2 NDVI and surface reflectance time series with 10 m spatial resolution, which is convenient for landcover classification and many other types of research.

In the practice of interpolating near-surface soil moisture measured by a wireless sensor network (WSN) grid, traditional Kriging methods with auxiliary variables, such as Co-kriging and Kriging with external drift (KED), cannot achieve satisfactory results because of the heterogeneity of soil moisture and its low correlation with the auxiliary variables. This study developed an Extended Kriging method to interpolate with the aid of remote sensing images. The underlying idea is to extend the traditional Kriging by introducing spectral variables, and operating on spatial and spectral combined space. The algorithm has been applied to WSN-measured soil moisture data in HiWATER campaign to generate daily maps from 10 June to 15 July 2012. For comparison, three traditional Kriging methods are applied: Ordinary Kriging (OK), which used WSN data only, Co-kriging and KED, both of which integrated remote sensing data as covariate. Visual inspections indicate that the result from Extended Kriging shows more spatial details than that of OK, Co-kriging, and KED. The Root Mean Square Error (RMSE) of Extended Kriging was found to be the smallest among the four interpolation results. This indicates that the proposed method has advantages in combining remote sensing information and ground measurements in soil moisture interpolation.

Atmospheric fine particulate matter (particulate matter with an aerodynamic diameter ≤ 2.5 μm in ambient air; PM _2.5 ) is a major pollutant causing regional air pollution and harm to human health. To monitor PM _2.5 , Chinese industry authorities use data from a small number of fixed stations with uneven distribution. Pollution control is promoted through assessment and enforcement, which can include industry-wide shutdown measures harmful to local economic development. The objective of this study was to establish a fine particulate matter network (FPMN) of sensors based on the Internet of Things with low cost, high spatiotemporal resolution, flexible distribution points, large numbers, and high collection frequency. The FPMN-derived data, together with other multisource environment-related data, could be used to track and locate atmospheric pollutants and selectively identify source. This study adopted Chizhou, China as the research object. Specifically, the work included designing the FPMN, selecting locations for sensor placement on the basis of local weather, terrain, and land use, and using software/hardware collaborative calibration technology to ensure consistency between FPMN-derived data and National Control Station(NCS) data. The analysis revealed that FPMN data effectively reconstructed a reliable regional field of PM _2.5 concentration with improved spatiotemporal accuracy. The research results will have great importance regarding the traceability of sources of PM _2.5 pollution, analysis of pollution causes and transboundary pollution, optimization and adjustment of industrial layouts, and differentiation of the temporal and spatial control of pollution. Ultimately, the FPMN could directly support management and decision-making processes of local governments in relation to the environment and industry.