Ten Lectures on Wavelets

Frames were introduced by Duffin and Schaeffer (1952), in the context of non-harmonic Fourier series (i.e., expansions of functions in L 2([0, 1]) in complex exponentials exp( i ? n x), where ? n ? 2 ? n); they are also reviewed in Young (1980). We review here their definition and some of their properties.
A family of functions ( ? j) j ?J in a Hilbert space
is called a frame if there exist A > 0, B < ? so that, for all f in
,
We call A and B the frame bounds.
If the two frame bounds are equal, A = B, then I will call the frame a tight frame. In a tight frame we have, for all f
,
which implies, by the polarization identity, [3]
or
at least in the weak sense. Formula (3.2.2) is very reminiscent of the expansion of f into an orthonormal basis, but it is important to realize that frames, even tight frames, are not orthonormal bases, as illustrated by the following finite-dimensional example.
Take
, e 1 = (0,1),
,
. (See Figure 3.1.) For any v = ( v 1, v 2) in
, we have

It follows that { e 1, e 2, e 3} is...