Skip to Main content Skip to Navigation
Journal articles

Improving k-nearest neighbor approaches for density-based pixel clustering in hyperspectral remote sensing images

Abstract : We investigated nearest-neighbor density-based clustering for hyperspectral image analysis. Four existing techniques were considered that rely on a K-nearest neighbor (KNN) graph to estimate local density and to propagate labels through algorithm-specific labeling decisions. We first improved two of these techniques, a KNN variant of the density peaks clustering method DPC, and a weighted-mode variant of KNNCLUST, so the four methods use the same input KNN graph and only differ by their labeling rules. We propose two regularization schemes for hyperspectral image analysis: (i) a graph regularization based on mutual nearest neighbors (MNN) prior to clustering to improve cluster discovery in high dimensions; (ii) a spatial regularization to account for correlation between neighboring pixels. We demonstrate the relevance of the proposed methods on synthetic data and hyperspectral images, and show they achieve superior overall performances in most cases, outperforming the state-of-the-art methods by up to 20% in kappa index on real hyperspectral images. © MDPI AG. All rights reserved.
Document type :
Journal articles
Complete list of metadata

https://hal.archives-ouvertes.fr/hal-03040362
Contributor : Admin Hal Ur1 Connect in order to contact the contributor
Submitted on : Friday, May 28, 2021 - 11:17:11 AM
Last modification on : Friday, October 22, 2021 - 3:04:10 AM
Long-term archiving on: : Sunday, August 29, 2021 - 6:51:00 PM

File

remotesensing-12-03745-v2.pdf
Publisher files allowed on an open archive

Identifiers

Citation

Claude Cariou, S.L. Moan, K. Chehdi. Improving k-nearest neighbor approaches for density-based pixel clustering in hyperspectral remote sensing images. Remote Sensing, MDPI, 2020, 12 (22), pp.1-29. ⟨10.3390/rs12223745⟩. ⟨hal-03040362⟩

Share

Metrics

Record views

66

Files downloads

56