The Yellow River occupies a pivotal strategic position in the development and economic construction of China. Moreover, grasping dynamics change long-term vegetation cover predicting future trends Basin could provide an empirical foundation for improved ecological protection soil water conservation initiatives. This study uses statistical methods such as Dimidiate pixel model, linear regression, Moran's index, coefficient variation to conduct spatio-temporal analysis coverage Basin. Hurst...

Semantic segmentation has emerged as a mainstream method in very-high-resolution remote sensing land-use/land-cover applications. In this paper, we first review the state-of-the-art semantic models both computer vision and fields. Subsequently, introduce two frameworks: SNFCN SDFCN, of which contain deep fully convolutional networks with shortcut blocks. We adopt an overlay strategy postprocessing method. Based on our frameworks, conducted experiments online ISPRS datasets: Vaihingen...

Change detection is of great significance in remote sensing. The advent high-resolution sensing images has greatly increased our ability to monitor land use and cover changes from space. At the same time, present a new challenge over other satellite systems, which time-consuming tiresome manual procedures must be needed identify changes. In recent years, deep learning (DL) been widely used fields natural image target detection, speech recognition, face etc., achieved success. Some scholars...

Object detection on very-high-resolution (VHR) remote sensing imagery has attracted a lot of attention in the field image automatic interpretation. Region-based convolutional neural networks (CNNs) have been vastly promoted this domain, which first generate candidate regions and then accurately classify locate objects existing these regions. However, overlarge images, complex backgrounds uneven size quantity distribution training samples make tasks more challenging, especially for small...

Over the last decade, object-based image classification (OBIC) has become a mainstream method in remote sensing land-use/land-cover applications. Many supervised methods have been proposed OBIC framework. However, most did not use deep learning methods. In this paper, new deep-learning-based framework is introduced. First, we segment original into objects by graph-based minimal-spanning-tree segmentation algorithm. Second, extract spectral, spatial, and texture features for each object. Then...

Scene classification, aiming to identify the land-cover categories of remotely sensed image patches, is now a fundamental task in remote sensing analysis field. Deep-learning-model-based algorithms are widely applied scene classification and achieve remarkable performance, but these high-level methods computationally expensive time-consuming. Consequently this paper, we introduce knowledge distillation framework, currently mainstream model compression method, into improve performance smaller...

Previously, generative adversarial networks (GAN) have been widely applied on super resolution reconstruction (SRR) methods, which turn low-resolution (LR) images into high-resolution (HR) ones. However, as these methods recover high frequency information with what they observed from the other images, tend to produce artifacts when processing unfamiliar images. Optical satellite remote sensing are of a far more complicated scene than natural Therefore, applying previous especially...

In this study, the addition of sulfamethazine (SMT) to landfill refuse decreased nitrogen intermediates (e.g. N2O and NO) dinitrogen (N2) gas fluxes <0.5 μg-N/kg-refuse·h-1, while N2 flux were at ~1.5 5.0 μg-N/kg-refuse·h-1 respectively in samples which oxytetracycline (OTC) had been added. The ARG (antibiotic resistance gene) levels increased tenfold after long-term exposure antibiotics, followed by a fourfold increase flux, but SMT-amended with largest resistome facilitated denitrification...

Semantic segmentation is a fundamental task in remote sensing image analysis (RSIA). Fully convolutional networks (FCNs) have achieved state-of-the-art performance the of semantic natural scene images. However, due to distinctive differences between images and remotely-sensed (RS) images, FCN-based methods from field computer vision cannot achieve promising performances on RS without modifications. In previous work, we proposed an framework SDFCNv1, combined with majority voting...

In recent years, remote sensing techniques such as satellite and drone-based imaging have been used to monitor Pine Wilt Disease (PWD), a widespread forest disease that causes the death of pine species. Researchers explored use imagery deep learning algorithms improve accuracy PWD detection at single-tree level. This study introduces novel framework for combines high-resolution RGB drone with free-access Sentinel-2 multi-spectral imagery. The proposed approach includes an PWD-infected tree...

Quinoline is biodegradable under anaerobic conditions, but information about the degradation kinetics and involved microorganisms scarce. Here, dynamics of a quinoline-degrading bacterial consortium were studied in anoxic batch cultures containing nitrate. The removed 83.5% quinoline during first 80 hours, which dominated by denitrification, then switched to methanogenesis when nitrogen oxyanions depleted. Time-resolved community analysis using pyrosequencing revealed that denitrifiying...

Satellite video single object tracking has attracted wide attention. The development of remote sensing platforms for earth observation technologies makes it increasingly convenient to acquire high-resolution satellite videos, which greatly accelerates ground target tracking. However, overlarge images with small size, high similarity among multiple moving targets, and poor distinguishability between the objects background make this task most challenging. To solve these problems, a deep...

Object-based image classification (OBIC) on very-high-resolution (VHR) remote sensing (RS) images is utilized in a wide range of applications. Nowadays, many existing OBIC methods only focus features each object itself, neglecting the contextual information among adjacent objects and resulting low accuracy. Inspired by spectral graph theory, we construct structure from generated VHR RS propose an framework based truncated sparse singular value decomposition convolutional network (GCN) model,...

Auto-extraction of convective clouds is great significance. Convective often bring heavy rain, strong winds, and other disastrous weather. Early warning convection can effectively reduce loss. Using remote sensing images, we get large-scale cloud information, which provides many effective methods for detection. In this letter, proposed a novel method to extract clouds. We introduce deep network using only <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML"...

Understanding complex urban systems necessitates untangling the relationships between diverse elements such as population, infrastructure, and socioeconomic activities. Scaling laws are basic but effective rules for evaluating a city's internal growth logic assessing its efficiency by investigating whether indicators scale with population. To date, only limited research has empirically explored scaling relations variables of mobility in mega-cities at an intra-urban few meters. Using...

Object-based image classification (OBIC) is presented to overcome the drawbacks of pixel-based (PBIC) when very-high-resolution (VHR) imagery classified. However, most methods in OBIC are dealing with 1D hand-crafted features extracted from segmented objects (superpixels). To extract 2D deep superpixels, a new framework introduced this letter by using convolutional neural networks (CNNs). We first analyze different mask policies superpixels and design two architectures networks. Then, we...

Space-filling designs based on orthogonal arrays are attractive for computer experiments they can be easily generated with desirable low-dimensional stratification properties. Nonetheless, it is not very clear how behave and to construct good such under other space-filling criteria. In this paper, we justify array-based a broad class of criteria, which include commonly used distance-, orthogonality- discrepancy-based measures. To identify even better properties, partition into classes by...

As a required step in optical remote sensing applications, image matching identifies correspondences to estimate the relationship between two images. To address this task, feature-based algorithms, such as Scale-Invariant Feature Transform (SIFT), use detectors identify keypoints and then apply descriptors represent these feature vectors. Thereby, vectors from different images are matched by Euclidean distance produce points correspondences. Deep learning networks widely used design of due...

Object detection has attracted a lot of attention in the field image automatic interpretation. Detectors based on convolution neural networks (CNNs) applied natural scene images encode results with horizontal bounding boxes (HBBs), which can not accurately calibrate position and shape arbitrary-orientation objects remote sensing (RSIs). To solve these issues, we propose an object framework named multi-oriented rotation-equivariant network (MORE-Net) this letter. The MORE-Net consists...