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

The very high spatial resolution (VHR) remote sensing images have been anextremely valuable source for monitoring changes occurred on the earth surface.However, precisely detecting relevant changes in VHR images still remains achallenge, due to the complexity of the relationships among ground objects. Toaddress this limitation, a dual neighborhood hypergraph neural network isproposed in this article, which combines the multiscale superpixel segmentationand hypergraph convolution to model and exploit the complex relationships.First, the bi-temporal image pairs are segmented under two scales and fed to apre-trained U-net to obtain node features by treating each object under thefine scale as a node. The dual neighborhood is then defined using thefather-child and adjacent relationships of the segmented objects to constructthe hypergraph, which permits models to represent the higher-order structuredinformation far more complex than just pairwise relationships. The hypergraphconvolutions are conducted on the constructed hypergraph to propagate the labelinformation from a small amount of labeled nodes to the other unlabeled ones bythe node-edge-node transform. Moreover, to alleviate the problem of imbalancedsample, the focal loss function is adopted to train the hypergraph neuralnetwork. The experimental results on optical, SAR and heterogeneous optical/SARdata sets demonstrate that the proposed method comprises better effectivenessand robustness compared to many state-of-the-art methods.

A Dual Neighborhood Hypergraph Neural Network for Change Detection in VHR Remote Sensing Images