Wavelet De-Noising of Hyperspectral Data

Abstract

This paper presents a method for noise removal from hyperspectral data using wavelets. Normally, data collected in the field will contain several types of noise. One type is the small amount of noise that is superimposed along the signal. This noise will not degrade the characteristic shape of the signal, but will add a small systematic disturbance to it. This noise is usually caused by the physical instrument or field environment conditions. This noise, although small, can be troublesome in interpolating data. The natural growth in the understanding of wavelet applications now affords the capability to remove this noise without degrading the signal. A wavelet transform was applied to hyperspectral percent reflectance - wavelength data sets. The resulting power spectrum was filtered so that specific wavelet coefficients were removed. An inverse wavelet transform was then applied to this filtered spectrum to obtain a noise free data set.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1996
Accession Number
ADA355101

Entities

People

  • E. Simental
  • T. Evans

Tags

Communities of Interest

  • Space

DTIC Thesaurus Topics

  • Camouflage
  • Coefficients
  • Computer Programs
  • Crystal Structure
  • Data Sets
  • Detection
  • Digital Images
  • Land Warfare
  • Mathematical Models
  • Mathematics
  • Measurement
  • Noise
  • Power Spectra
  • Reflectance
  • Remote Sensing
  • Sine Waves
  • Wavelet Transforms

Readers

  • Atmospheric Science / Meteorology, specifically Wind Wave Turbulence.
  • Image Processing and Computer Vision.