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...

Täydet tiedot

Bibliografiset tiedot

Päätekijät:	Saragadam, Vishwanath, Sankaranarayanan, Aswin C.
Aineistotyyppi:	Artikkeli
Julkaistu:	2019
Aiheet:	Image and Video Processing Computer Vision and Pattern Recognition Machine Learning
Linkit:	https://arxiv.org/abs/1905.04815 http://arxiv.org/abs/1905.04815
Tagit:	Lisää tagi Ei tageja, Lisää ensimmäinen tagi!