Abstract

We have developed an adaptive spectral imaging classifier that directly spectrally-classifies every location in a scene and adaptively adjusts itself to facilitate this classification process. The net result is classification error rates that are multiple orders-of-magnitude below those of conventional instruments or even static computational/compressive sensing approaches. I will present our latest results and demonstrate quantitative agreement with expected performance across a wide range of operational parameters. Article not available.

© 2015 OSA

PDF Article | Presentation Video

Supplementary Material (1)

Presentation Video