Noise-signal index threshold: a new noise-reduction technique for generation of reference spectra and efficient hyperspectral image classification

Abstract

Reference spectra of terrestrial targets are usually collected using field spectro-radiometers for mineral abundance mapping and target detection. These spectra often have noise that masks characteristic absorption and reflection features and affects the efficiency of material mapping. This work aims at obtaining an empirical technique for reduction of high-frequency noise from field spectra. The proposed noise correction technique uses a ‘normalized’ measure R_n , where R_n  = (L_n  − F_n )/L_n for each band (n) calculated from field and laboratory spectra of test material, with F_n and L_n being the depth of the absorption feature in field and laboratory spectra, respectively. On the basis of the assumption of the constancy of this ratio in neighbouring bands, an empirical algorithm that approximates the ratio R_n of a noisy band to the corrected ratio of an adjacent band is used to obtain the noise-corrected field spectra. The classification accuracy increases significantly when noise reduced field spectra are used as reference spectra.

Keywords:

• reflectance spectra
• noise
• empirical correction
• classification
• hyperspectral remote sensing

Acknowledgements

The authors are thankful to Professor S.S. Major, IIT, Bombay and Dr. R. N. Sahoo, IARI, New Delhi for instrumentation support. Constructive criticism from two unknown reviewers is highly appreciated for their help in improving the quality of the manuscript. D.R. acknowledges the Department of Science and Technology, Government of India for financial support.