On-the-fly spectral unmixing for real-time hyperspectral data analysis
Abstract
This paper presents an online linear unmixing method leveraging Kalman filtering for real-time analysis of hyperspectral data. Unlike traditional methods that require processing of the entire data set, the proposed approach sequentially updates pure spectra estimates as new data is acquired, that is, on a spectrum-by-spectrum basis, thereby significantly reducing computational cost. Experiments conducted on synthetic and real Raman data sets demonstrate that the proposed method achieves a favorable trade-off between unmixing accuracy and computational efficiency, making it suitable for real-time hyperspectral imaging applications.
| Origin | Files produced by the author(s) |
|---|