REVIVE preliminary study 2: band selection on real InGaAs hyperspectral images of construction materials (900–1.700 nm)

Our first study (doi:10.5281/zenodo.24280020) used library point spectra. This one checks the same two claims on pixel spectra from a real InGaAs hyperspectral imaging sensor, the detector type of the Goldeye G-130 VSWIR: (1) that wavelengths above 1.000 nm separate building materials that a silicon sensor cannot, and (2) that a few well-chosen bands keep most of that information. On the authors’ test set, five bands chosen over 900–1.700 nm reach 71 % accuracy across six materials, against 39 % with five bands limited to the silicon range (900–1.000 nm) and 80 % with all 224 bands. Source data: “Data for: A near-infrared hyperspectral data recognition algorithm for construction waste and its experimental study”, Mendeley Data, doi:10.17632/x7kbw53rxk.1 (CC BY-NC 3.0), used for non-commercial research with attribution and not redistributed here. This record contains only our note, code and results.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-19
DOI
https://doi.org/10.5281/zenodo.24365293
Primary Topic
Remote-Sensing Image Classification
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

REVIVE preliminary study 2: band selection on real InGaAs hyperspectral images of construction materials (900–1.700 nm)

Alberto Jose Arenillas
Zenodo (CERN European Organization for Nuclear Research)
Remote-Sensing Image Classification
article

REVIVE preliminary study 2: band selection on real InGaAs hyperspectral images of construction materials (900–1.700 nm)

Alberto Jose Arenillas
article en

Abstract

Our first study (doi:10.5281/zenodo.24280020) used library point spectra. This one checks the same two claims on pixel spectra from a real InGaAs hyperspectral imaging sensor, the detector type of the Goldeye G-130 VSWIR: (1) that wavelengths above 1.000 nm separate building materials that a silicon sensor cannot, and (2) that a few well-chosen bands keep most of that information. On the authors’ test set, five bands chosen over 900–1.700 nm reach 71 % accuracy across six materials, against 39 % with five bands limited to the silicon range (900–1.000 nm) and 80 % with all 224 bands. Source data: “Data for: A near-infrared hyperspectral data recognition algorithm for construction waste and its experimental study”, Mendeley Data, doi:10.17632/x7kbw53rxk.1 (CC BY-NC 3.0), used for non-commercial research with attribution and not redistributed here. This record contains only our note, code and results.

Zenodo (CERN European Organization for Nuclear Research)
Remote-Sensing Image Classification
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon