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

Trees inside cities are important for the urban microclimate, contributingpositively to the physical and mental health of the urban dwellers. Despitetheir importance, often only limited information about city trees is available.Therefore in this paper, we propose a method for mapping urban trees inhigh-resolution aerial imagery using limited datasets and deep learning. Deeplearning has become best-practice for this task, however, existing approachesrely on large and accurately labelled training datasets, which can be difficultand expensive to obtain. However, often noisy and incomplete data may beavailable that can be combined and utilized to solve more difficult tasks thanthose datasets were intended for. This paper studies how to combine accuratepoint labels of urban trees along streets with crowd-sourced annotations froman open geographic database to delineate city trees in remote sensing images, atask which is challenging even for humans. To that end, we perform semanticsegmentation of very high resolution aerial imagery using a fully convolutionalneural network. The main challenge is that our segmentation maps are sparselyannotated and incomplete. Small areas around the point labels of the streettrees coming from official and crowd-sourced data are marked as foregroundclass. Crowd-sourced annotations of streets, buildings, etc. define thebackground class. Since the tree data is incomplete, we introduce a masking toavoid class confusion. Our experiments in Hamburg, Germany, showed that thesystem is able to produce tree cover maps, not limited to trees along streets,without providing tree delineations. We evaluated the method on manuallylabelled trees and show that performance drastically deteriorates if the opengeographic database is not used.

Predicting urban tree cover from incomplete point labels and limited background information