Programmable Spectrometry -- Per-pixel Classification of Materials using Learned Spectral Filters
Many materials have distinct spectral profiles. This facilitates estimation of the material composition of a scene at each pixel by first acquiring its hyperspectral image, and subsequently filtering it using a bank of spectral profiles. This process is inherently wasteful since only a set of linear...
Enregistré dans:
| Auteurs principaux: | , |
|---|---|
| Format: | Article |
| Publié: |
2019
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/1905.04815 http://arxiv.org/abs/1905.04815 |
| Tags: |
Pas de tags, Soyez le premier à ajouter un tag!
|