Digital Signal Processing Reference
In-Depth Information
; Example of Starlet Transform
; Read an image of a Galaxy
n = float( read_png('ngc2997.png'))
; Compute its starlet wavelet transform with six scales
W = star2d(n ,nscale=6)
; visualize the different wavelet scales
for j=0,5 do begin tvscl, W[*,*,j] & wait, 2 & end
; Dynamic Range Compression
; Apply the dynamic range compression to the comet image
; To read the FITS format, the IDL Astronomical User Library must
; be installed (http://idlastro.gsfc.nasa.gov).
h = readfits('HaleBopp256.fits')
; Apply the dynamic range compression of h, using five scales,
; visualize it.
tvscl,
star2d_drc(h,nscale=5, eps=0.1)
; Apply the dynamic range compression of the ophthalmic image, using
; five scales, and visualize it.
o = float( read_png('opthalmic.png'))
tvscl, star2d_drc(o,nscale=5)
3.9 SUMMARY
In this chapter, we have focused on effective and efficient discrete redundant
wavelet transform algorithms. Effectiveness is important for such goals as denois-
ing (restoration in general), faint feature detection, or pattern analysis, so scien-
tific, medical, or forensic imaging applications are paramount. Efficiency is primarily
computational, but we also pay attention to storage needs.
