English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Reservoirs are fundamental infrastructures for the management of waterresources. Constructions around them can negatively impact their quality. Suchunauthorized constructions can be monitored by land cover mapping (LCM) remotesensing (RS) images. In this paper, we develop a new approach based on DL andimage processing techniques for man-made object segmentation around thereservoirs. In order to segment man-made objects around the reservoirs in anend-to-end procedure, segmenting reservoirs and identifying the region ofinterest (RoI) around them are essential. In the proposed two-phase workflow,the reservoir is initially segmented using a DL model. A post-processing stageis proposed to remove errors such as floating vegetation. Next, the RoI aroundthe reservoir (RoIaR) is identified using the proposed image processingtechniques. Finally, the man-made objects in the RoIaR are segmented using a DLarchitecture. We trained the proposed workflow using collected Google Earth(GE) images of eight reservoirs in Brazil over two different years. TheU-Net-based and SegNet-based architectures are trained to segment thereservoirs. To segment man-made objects in the RoIaR, we trained and evaluatedfour possible architectures, U-Net, FPN, LinkNet, and PSPNet. Although thecollected data has a high diversity (for example, they belong to differentstates, seasons, resolutions, etc.), we achieved good performances in bothphases. Furthermore, applying the proposed post-processing to the output ofreservoir segmentation improves the precision in all studied reservoirs excepttwo cases. We validated the prepared workflow with a reservoir dataset outsidethe training reservoirs. The results show high generalization ability of theprepared workflow.

An End-to-End Two-Phase Deep Learning-Based workflow to Segment Man-made Objects Around Reservoirs