Cloud Deep Networks for Hyperspectral Image Analysis

Haut, Juan M.; Gallardo, Jose Antonio; Paoletti, Mercedes E.; Cavallaro, Gabriele; Plaza, Javier; Plaza, Antonio; Riedel, Morris

doi:10.1109/tgrs.2019.2929731

Cited by 27 publications

(14 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yet, locally hosted software options are unlikely to be the future for hyperspectral data processing. Hyperspectral data fits within the category of "big geo data" (Krämer and Senner, 2015) and is, therefore, better suited to scalable and distributed cloud processing rather than local computing capabilities (Wilson et al, 2018;Haut et al, 2019). Although cloud-based high-performance computing (HPC) is not a new concept (e.g., Plaza et al, 2011), its intersection with hyperspectral data for environmental analyses-particularly in aquatic environments-is in its infancy and remains an area for considerable future growth.…”

Section: Peering Into the Abyssmentioning

confidence: 99%

Living up to the Hype of Hyperspectral Aquatic Remote Sensing: Science, Resources and Outlook

Dierssen

Ackleson

Joyce

et al. 2021

Front. Environ. Sci.

View full text Add to dashboard Cite

Intensifying pressure on global aquatic resources and services due to population growth and climate change is inspiring new surveying technologies to provide science-based information in support of management and policy strategies. One area of rapid development is hyperspectral remote sensing: imaging across the full spectrum of visible and infrared light. Hyperspectral imagery contains more environmentally meaningful information than panchromatic or multispectral imagery and is poised to provide new applications relevant to society, including assessments of aquatic biodiversity, habitats, water quality, and natural and anthropogenic hazards. To aid in these advances, we provide resources relevant to hyperspectral remote sensing in terms of providing the latest reviews, databases, and software available for practitioners in the field. We highlight recent advances in sensor design, modes of deployment, and image analysis techniques that are becoming more widely available to environmental researchers and resource managers alike. Systems recently deployed on space- and airborne platforms are presented, as well as future missions and advances in unoccupied aerial systems (UAS) and autonomous in-water survey methods. These systems will greatly enhance the ability to collect interdisciplinary observations on-demand and in previously inaccessible environments. Looking forward, advances in sensor miniaturization are discussed alongside the incorporation of citizen science, moving toward open and FAIR (findable, accessible, interoperable, and reusable) data. Advances in machine learning and cloud computing allow for exploitation of the full electromagnetic spectrum, and better bridging across the larger scientific community that also includes biogeochemical modelers and climate scientists. These advances will place sophisticated remote sensing capabilities into the hands of individual users and provide on-demand imagery tailored to research and management requirements, as well as provide critical input to marine and climate forecasting systems. The next decade of hyperspectral aquatic remote sensing is on the cusp of revolutionizing the way we assess and monitor aquatic environments and detect changes relevant to global communities.

show abstract

Section: Peering Into the Abyssmentioning

confidence: 99%

Living up to the Hype of Hyperspectral Aquatic Remote Sensing: Science, Resources and Outlook

Dierssen

Ackleson

Joyce

et al. 2021

Front. Environ. Sci.

View full text Add to dashboard Cite

show abstract

“…The situation is steadily improving with the great availability of deep neural network (DNN) frameworks (Robinson et al, 2019a), within well-documented libraries (Zhu et al, 2017). It is accompanied by the set up of multiple High Performance Computing infrastructures (HPC) and libraries, mandatory so as to upscale proposed workflows, swallow the amount of geospatial images, and provide maps in decent times (Chi et al, 2016, Haut et al, 2019. Infrastructures are located either in universities/institutes and national data centers, or are now directly provided by/rent to private companies (Google, Amazon Web Services, Microsoft Azure).…”

Section: Paradigm Change In Land-cover Mappingmentioning

confidence: 99%

Current Challenges in Operational Very High Resolution Land-Cover Mapping

Mallet

Bris

2020

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Abstract. Many land-cover products have been made available for a large range of end-users over the last ten years, even at global scales. In particular, remote sensing data analysis has proved to be the most feasible solution for automation purposes, at multiple spatial scales. However, current solutions are not sufficient for designing better products, adapted to real-case applications, operational constraints, and the generation of services, built upon these core layers. In this paper, we review the main requirements and the recent changes in remote sensing for the specific case of very high resolution land-cover mapping. We also comment current and evaluate challenges for the optimal exploitation of Earth Observation images with the aim of automatically generating maps tailored to specific end-users’ needs. We advocate for more challenging large-scale benchmarks and for human-in-the-loop solutions.

show abstract

“…HSI classification algorithms generally map nicely on HPC systems [14], with several HPC approaches (from coarse-grained and fine-grained parallelism techniques to complex distributed environments) already successfully implemented. For instance, distributed approaches based on commodity (homogeneous and heterogeneous) clusters [15,16], grid computing [17,18], and cloud computing techniques [19][20][21][22] provided good results in terms of reducing execution times, enabling an efficient processing of massive data sets on the ground segment. However, dedicated resources such as massively parallel clusters and networks of computers are generally expensive and hard to maintain, being cloud computing a more appropriate and cheaper approach (in addition to a fully distributed solution) due to its "service-oriented" computing and "pay-per-use" model [23].…”

Section: Introductionmentioning

confidence: 99%

A New GPU Implementation of Support Vector Machines for Fast Hyperspectral Image Classification

et al. 2020

Self Cite

View full text Add to dashboard Cite

The storage and processing of remotely sensed hyperspectral images (HSIs) is facing important challenges due to the computational requirements involved in the analysis of these images, characterized by continuous and narrow spectral channels. Although HSIs offer many opportunities for accurately modeling and mapping the surface of the Earth in a wide range of applications, they comprise massive data cubes. These huge amounts of data impose important requirements from the storage and processing points of view. The support vector machine (SVM) has been one of the most powerful machine learning classifiers, able to process HSI data without applying previous feature extraction steps, exhibiting a robust behaviour with high dimensional data and obtaining high classification accuracies. Nevertheless, the training and prediction stages of this supervised classifier are very time-consuming, especially for large and complex problems that require an intensive use of memory and computational resources. This paper develops a new, highly efficient implementation of SVMs that exploits the high computational power of graphics processing units (GPUs) to reduce the execution time by massively parallelizing the operations of the algorithm while performing efficient memory management during data-reading and writing instructions. Our experiments, conducted over different HSI benchmarks, demonstrate the efficiency of our GPU implementation.

show abstract

Cloud Deep Networks for Hyperspectral Image Analysis

Cited by 27 publications

References 52 publications

Living up to the Hype of Hyperspectral Aquatic Remote Sensing: Science, Resources and Outlook

Living up to the Hype of Hyperspectral Aquatic Remote Sensing: Science, Resources and Outlook

Current Challenges in Operational Very High Resolution Land-Cover Mapping

A New GPU Implementation of Support Vector Machines for Fast Hyperspectral Image Classification

Contact Info

Product

Resources

About