|
Best methods use direct 2D methods starting with 2 by 2 or similar pixel blocks. However one can compose in several ways 1D transforms. It is NOT best to first transform in x and then transform in y. Rather ...
|
A nice method uses two separate but INTERMIXED 1D transforms
-
Image filtered along the x dimension
-
Downsample by 2 along x
-
Filter along the y dimension
-
Downsample by 2 along y
|
|
Recursively transform the average signal (depending on required ratio)
|
|
We have matrix of coefficients (average signal and detail signals of each scale); no compression has been accomplished yet; in fact there has been an increase in amount of storage required
|
