13.4 3-D DIGITAL FILTERS
A first-quadrant 3-D recursive filter can be represented by the equation
(13.15)
![]()
where X ≡ X (z1, z2, z3), Y ≡ Y (z1, z2, z3), and b(0, 0, 0) = 0. In the above equation, h is the height of the filter window, w is the width of the filter window, and f is the depth of the filter window, which is the number of frames used to collect the filter input samples.
Assuming progressive raster-scanned data, the term
represents one time-step delay through the use of a single register as a storage element. The term
represents W time-step delays where W is the frame width. The term
represents HW time-step delays where H is the frame height. These delays could be implemented as memory address locators for use by the software threads, or they could be implemented as actual hardware buffers for use by the software systolic array processing elements.
The above equation can be written in a hierarchical way as
where the terms F2(k3) and G2(k3) are two 2-D IIR filter ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access