Base implementation of psImageCut/psImageRadialCut

[robert.desonia@…]

In a polycon a couple of weeks before rel2 shipment, Eugene stated he had an
example piece of code for psImageCut.

Could we get this code or should we move ahead without it?

The thing in question is how exactly one would group the pixels within a w
width about an arbitrary line cut for psImageCut. Is the rectangular region
arbitrarily positioned over the image split into p by w segments (where p is
the pixel width of the image?) and all pixel values within this segment used
given the psStats struct? Is the center of this segment the start at the
start of the line, or 0.5*p beyond it? (etc., etc., etc.)

The identification of pixels under this segmented rectangle could be a costly
operation; do you any suggestions for a well known algorithm in which you'd
like to see implemented?

-rdd

[eugene]

I will write this up and get it to you within ~1-2 days. the details are a
little more subtle than I originally was considering. For example, the pixels
passed to the psImageStats (or psVectorStats) are quantized; you can only pass
an integer number of pixels. What do you do with the fractional pixels. There
are a few ways to approach this, but I need to write it down carefully.

[eugene]

this has been added to the ADD, (v.05) along with some changes to the
implementation in the SDRS (v.06)

[Paul Price]

Should be fixed in SDRS-07 and ADD-06 (7 September 2004).