Spectral observations along the spectrum in many narrow spectral bands through hyperspectral imaging provides valuable information towards material and object recognition, which can be consider as a classification task. Most of existing studies research efforts are following conventional pattern recognition paradigm, is based on construction complex handcrafted features. However, it rarely known features important for problem at hand. In contrast to these approaches, we propose deep learning...

The automated man-made object detection and building extraction from single satellite images is, still, one of the most challenging tasks for various urban planning monitoring engineering applications. To this end, in paper we propose an framework very high resolution remote sensing data based on deep convolutional neural networks. core developed method is a supervised classification procedure employing large training dataset. An MRF model then responsible obtaining optimal labels regarding...

In this article, we present a deep multitask learning framework able to couple semantic segmentation and change detection using fully convolutional long short-term memory (LSTM) networks. particular, UNet-like architecture (L-UNet) that models the temporal relationship of spatial feature representations integrated LSTM blocks on top every encoding level. way, network is capture vectors in all levels without need downsample or flatten them, forming an end-to-end trainable framework. Moreover,...

The advent of multitemporal high resolution data, like the Copernicus Sentinel-2, has enhanced significantly potential monitoring earth's surface and environmental dynamics. In this paper, we present a novel deep learning framework for urban change detection which combines state-of-the-art fully convolutional networks (similar to U-Net) feature representation powerful recurrent (such as LSTMs) temporal modeling. We report our results on recently publicly available bi-temporal Onera Satellite...

Plastic debris in the global ocean is considered an important issue with severe implications for human health and marine ecosystems. Here, we exploited high-resolution multispectral satellite observations over Bay Islands Gulf of Honduras, period 2014-2019, to investigate capability sensors detecting plastic debris. We verified findings situ data, recorded spectral characteristics floating litter, identified trajectories sources. The results showed that originating from Guatemala’s Honduras’...

Abstract Marine litter is one of the most relevant pollution problems that our oceans are facing today. in a major threat to sustainable planet. Here, we provide comprehensive analysis cutting-edge solutions developed globally prevent, monitor and clean marine litter. Prevention this research includes only innovative prevent entering seas rather than interventions such as waste reduction recycling. On basis extensive search data compilation, reveals information dispersed across platforms not...

Currently, a significant amount of research is focused on detecting Marine Debris and assessing its spectral behaviour via remote sensing, ultimately aiming at new operational monitoring solutions. Here, we introduce Archive (MARIDA), as benchmark dataset for developing evaluating Machine Learning (ML) algorithms capable Debris. MARIDA the first based multispectral Sentinel-2 (S2) satellite data, which distinguishes from various marine features that co-exist, including Sargassum macroalgae,...

Despite the significant negative impact of marine pollution on ecosystem and humans, its automated detection tracking from broadly available satellite data is still a major challenge. In particular, most research development efforts focus one specific pollutant implementing, in cases, binary classification tasks, e.g., detect Plastics or no Plastics, target limited number classes, such as Oil Spill, Look-alikes Water. Moreover, developed algorithms tend to operate successfully only locally,...

In this paper, a novel recognition-driven variational framework, toward multiple building extraction from aerial and satellite images, is introduced. To end, competing shape priors are considered, addressed through an image segmentation approach that involves the use of data-driven term constrained prior models. The proposed framework extends previous approaches integration into level-set segmentation. particular, it estimates number buildings as well their pose observed data. Therefore, can...

In this paper an automated vehicle detection and traffic density estimation algorithm has been developed validated for very high resolution satellite video data. The is based on adaptive background procedure followed by a subtraction at every frame. performed through further mathematical morphology statistical analysis the computed connected components. estimated lower grid superimposed scene. particular, subregion number of detected vehicles calculated then entire road network...

Abstract. The indirect estimation of leaf area index (LAI) in large spatial scales is crucial for several environmental and agricultural applications. To this end, paper, we compare evaluate LAI vineyards from different UAV imaging datasets. In particular, canopy levels were estimated i.e., (i) hyperspectral data, (ii) 2D RGB orthophotomosaics (iii) 3D crop surface models. computed have been used to establish relationships with the measured (ground truth) vines Nemea, Greece. overall...

In this paper, the scientific outcomes of 2016 Data Fusion Contest organized by Image Analysis and Technical Committee IEEE Geoscience Remote Sensing Society are discussed. The was an open topic competition based on a multitemporal multimodal dataset, which included temporal pair very high resolution panchromatic multispectral Deimos-2 images video captured Iris camera on-board International Space Station. problems addressed techniques proposed participants to spanned across rather broad...

Although aerial image-based bathymetric mapping can provide, unlike acoustic or LiDAR (Light Detection and Ranging) sensors, both water depth visual information, refraction poses significant challenges for accurate estimation. In order to tackle this challenge, we propose an image correction methodology, which first exploits recent machine learning procedures that recover from dense point clouds then corrects on the original imaging dataset. This way, structure motion (SfM) multi-view stereo...

Abstract. The determination of accurate bathymetric information is a key element for near offshore activities, hydrological studies such as coastal engineering applications, sedimentary processes, hydrographic surveying well archaeological mapping and biological research. UAV imagery processed with Structure from Motion (SfM) Multi View Stereo (MVS) techniques can provide low-cost alternative to established shallow seabed offering the important visual information. Nevertheless, water...

The past few years have seen an accelerating integration of deep learning (DL) techniques into various remote sensing (RS) applications, highlighting their power to adapt and achieving unprecedented advancements. In the present review, we provide exhaustive exploration DL approaches proposed specifically for spatial downscaling RS imagery. A key contribution our work is presentation major architectural components models, metrics, data sets available this task as well construction a compact...

Abstract. In this paper we evaluated deep-learning frameworks based on Convolutional Neural Networks for the accurate classification of multispectral remote sensing data. Certain state-of-the-art models have been tested publicly available SAT-4 and SAT-6 high resolution satellite datasets. particular, performed benchmark included AlexNet, AlexNet-small VGG which had trained applied to both datasets exploiting all spectral information. Deep Belief Networks, Autoencoders other semi-supervised...

Deep learning architectures have received much attention in recent years demonstrating state-of-the-art performance several segmentation, classification and other computer vision tasks. Most of these deep networks are based on either convolutional or fully architectures. In this paper, we propose a novel object-based deep-learning framework for semantic segmentation very high-resolution satellite data. particular, exploit priors integrated into neural network by incorporating an anisotropic...

The determination of accurate bathymetric information is a key element for near offshore activities; hydrological studies, such as coastal engineering applications, sedimentary processes, hydrographic surveying, archaeological mapping and biological research. Through structure from motion (SfM) multi-view-stereo (MVS) techniques, aerial imagery can provide low-cost alternative compared to LiDAR (Light Detection Ranging) surveys, it offers additional important visual higher spatial...

Broccoli is an example of a high-value crop that requires delicate handling throughout the growing season and during its post-harvesting treatment. As broccoli heads can be easily damaged, they are still harvested by hand. Moreover, human scouting required to initially identify field segments where several plants have reached desired maturity level, such while in optimal condition. The aim this study was automate process using state-of-the-art Object Detection architectures trained on...

Savannah ecosystems face significant threats from land degradation, including woody vegetation encroachment. This study introduces a high-resolution method for mapping the fraction of savannah cover by integrating optical (Sentinel-2, S2), radar (Sentinel-1, S1), and auxiliary data. First, comprehensive training dataset fractional (FWC) samples was developed very imagery thousands manually annotated points. Shallow deep learning algorithms were utilised to generate classification masks, with...