Related Experiment Video
Updated: May 13, 2026
07:05
Applying Hyperspectral Reflectance Imaging to Investigate the Palettes and the Techniques of Painters
Published on: June 18, 2021
Separability between pedestrians in hyperspectral imagery.
Jared Herweg1, John Kerekes, Michael Eismann
1Air Force Institute of Technology, Dayton, Ohio 45433, USA. jxh6389@cis.rit.edu
Applied Optics
|February 26, 2013
Summary
Hyperspectral imaging (HSI) can detect pedestrians by analyzing spectral reflectance. Clothing, particularly shirts, provides the most distinct features for identification, outperforming skin reflectance.
Area of Science:
- Remote Sensing
- Spectroscopy
- Computer Vision
Background:
- Hyperspectral imaging (HSI) is increasingly adopted for novel remote sensing applications.
- Detecting and tracking human pedestrians presents unique imaging challenges.
Purpose of the Study:
- To explore the use of HSI for detecting and tracking human pedestrians.
- To assess the distinguishability of pedestrians based on spectral reflectance and noise.
Main Methods:
- Analyzed spectral separation of pedestrian material subregions (450-2500 nm).
- Evaluated spectral content combination from multiple subregions.
- Assessed separability using a unique hyperspectral image dataset.
Main Results:
- Skin reflectance showed minimal distinguishability between pedestrians.
- Clothing, especially shirts, offered the most salient distinguishing spectral features.
- Combining reflectance from multiple subregions significantly improved separability.
Conclusions:
- HSI is a viable technology for pedestrian detection and tracking.
- Clothing spectral signatures are key identifiers.
- Multi-spectral data fusion enhances pedestrian separability in HSI.
