Digital Signal Processing Reference
In-Depth Information
(a)
(b)
(c)
(d)
Figure 3.17. (a) Original “Einstein” image and (b) its noisy version with noise standard
deviation
σ
n
= 20. (c) Denoised image by the UWT (PSNR = 29.6 dB). (d) Denoised
image by the orthogonal DWT transform (PSNR = 27.8 dB).
3.8.2 Dynamic Range Compression Using the IDL Starlet Transform
In this section, we discuss IDL code. For support of Flexible Image Transport
System (FITS) format and other operations, the IDL Astronomical User Library
(http://idlastro.gsfc.nasa.gov) can be used. This is linked at the topic's Web address
(http://www.SparseSignalRecipes.info).
Because some features in an image may be hard to detect by the human eye due
to low contrast, we often process the image before visualization. Histogram equal-
ization is one of the most well known methods for contrast enhancement. Images
with a high dynamic range are also difficult to analyze. For example, astronomers
generally visualize their images using a logarithmic look-up table conversion.
