Challenges and limitations for cloud-based platforms and integration with AI algorithms for earth observation data analytics

Elsevier eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 513 - 526

Published: Jan. 1, 2025

Language: Английский

MMS-EF: A Multi-Scale Modular Extraction Framework for Enhancing Deep Learning Models in Remote Sensing

Land, Journal Year: 2024, Volume and Issue: 13(11), P. 1842 - 1842

Published: Nov. 5, 2024

The analysis of land cover using deep learning techniques plays a pivotal role in understanding use dynamics, which is crucial for management, urban planning, and cartography. However, due to the complexity remote sensing images, models face practical challenges preprocessing stage, such as incomplete extraction large-scale geographic features, loss fine details, misalignment issues image stitching. To address these issues, this paper introduces Multi-Scale Modular Extraction Framework (MMS-EF) specifically designed enhance applications. framework incorporates three key components: (1) multiscale overlapping segmentation module that captures comprehensive geographical information through multi-channel processing, ensuring integrity features; (2) feature fusion integrates local global facilitating seamless stitching improving classification accuracy; (3) detail enhancement refines small-scale enriching semantic imagery. Extensive experiments were conducted across various models, was validated on two public datasets. results demonstrate proposed approach effectively mitigates limitations traditional methods, significantly accuracy exhibiting strong adaptability different

Language: Английский

Integration of WebODM Platform as Photogrammetry Cloud Processing and GeoNode as Corporate Spatial Data Infrastructure to Support the Acceleration of Geospatial Data Communication Flow in PT. Waskita Karya Tbk.

IOP Conference Series Earth and Environmental Science, Journal Year: 2024, Volume and Issue: 1418(1), P. 012078 - 012078

Published: Dec. 1, 2024

Abstract This research explores the integration of two platforms, namely WebODM and GeoNode, with Geoserver as its connecting platform. platform is a solution to support acceleration geospatial data communication flow in PT. Waskita Karya Tbk. The main objective combine advantages Photogrammetry Cloud Processing owned by Spatial Data Infrastructure features thus creating an efficient integrated system for processing storing data. methodology involves user needs analysis, development, well testing performance evaluation implemented system. results indicate that GeoNode has successfully created environment allows users easily perform cloud-based photogrammetry processes efficiently store share With this system, can improve operational terms distribution, accelerate between various involved units. implication like be effective companies managing on large scale.

Language: Английский