Segmentation of hyperspectral images from functional kernel density estimation

Laurent Delsol (1) , Cécile Louchet (1)

1 MAPMO - Mathématiques - Analyse, Probabilités, Modélisation - Orléans

The processing of hyperspectral images, seen as functions that link each pixel to a curve, has become crucial, in remote sensing applications for instance. Here we tackle the problem of segmentation of such images, by carefully combining image processing tools and functional statistics, namely a Potts model and a likelihood term based on functional kernel density estimation in a Bayesian framework, and consider possible extensions.

Domaines: Autres [stat.ML]; Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Laurent Delsol, Cécile Louchet. Segmentation of hyperspectral images from functional kernel density estimation. International workshop on functional and operatorial statistics, Jun 2014, Stresa, Italy. pp.101-105. ⟨hal-01032419⟩