Hyperspectral Image Segmentation Using a New Spectral Unmixing-Based Binary Partition Tree Representation

Abstract : The Binary Partition Tree (BPT) is a hierarchical region-based representation of an image in a tree structure. BPT allows users to explore the image at different segmentation scales. Often, the tree is pruned to get a more compact representation and so the remaining nodes conform an optimal partition for some given task. Here, we propose a novel BPT construction approach and pruning strategy for hyperspectral images based on spectral unmixing concepts. Linear Spectral Unmixing (LSU) consists of finding the spectral signatures of the materials present in the image (endmembers) and their fractional abundances within each pixel. The proposed methodology exploits the local unmixing of the regions to find the partition achieving a global minimum reconstruction error. Results are presented on real hyperspectral data sets with different contexts and resolutions.
Complete list of metadatas

Cited literature [45 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01010430
Contributor : Miguel Angel Veganzones <>
Submitted on : Wednesday, November 25, 2015 - 2:47:46 PM
Last modification on : Wednesday, September 19, 2018 - 1:15:01 AM
Long-term archiving on : Saturday, April 29, 2017 - 3:47:36 AM

File

IEEE_TIP_Veganzones_final_onec...
Files produced by the author(s)

Identifiers

Citation

Miguel Angel Veganzones, Guillaume Tochon, Mauro Dalla-Mura, Antonio J. Plaza, Jocelyn Chanussot. Hyperspectral Image Segmentation Using a New Spectral Unmixing-Based Binary Partition Tree Representation. IEEE Transactions on Image Processing, Institute of Electrical and Electronics Engineers, 2014, 23 (8), pp.3574 - 3589. ⟨10.1109/TIP.2014.2329767⟩. ⟨hal-01010430⟩

Share

Metrics

Record views

465

Files downloads

573