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

In this paper, we introduce a framework for segmenting instances of a commonobject class by multiple active contour evolution over semantic segmentationmaps of images obtained through fully convolutional networks. The contourevolution is cast as an energy minimization problem, where the aggregate energyfunctional incorporates a data fit term, an explicit shape model, and accountsfor object overlap. Efficient solution neighborhood operators are proposed,enabling optimization through metaheuristics such as simulated annealing. Weinstantiate the proposed framework in the context of segmenting individualfallen stems from high-resolution aerial multispectral imagery. We validatedour approach on 3 real-world scenes of varying complexity. The test plots weresituated in regions of the Bavarian Forest National Park, Germany, whichsustained a heavy bark beetle infestation. Evaluations were performed on boththe polygon and line segment level, showing that the multi-contour segmentationcan achieve up to 0.93 precision and 0.82 recall. An improvement of up to 7percentage points (pp) in recall and 6 in precision compared to an iterativesample consensus line segment detection was achieved. Despite the simplicity ofthe applied shape parametrization, an explicit shape model incorporated intothe energy function improved the results by up to 4 pp of recall. Finally, weshow the importance of using a deep learning based semantic segmentation methodas the basis for individual stem detection. Our method is a step towardsincreased accessibility of automatic fallen tree mapping, due to higher costefficiency of aerial imagery acquisition compared to laser scanning. Theprecise fallen tree maps could be further used as a basis for plant and animalhabitat modeling, studies on carbon sequestration as well as soil quality inforest ecosystems.

Instance segmentation of fallen trees in aerial color infrared imagery  using active multi-contour evolution with fully convolutional network-based  intensity priors