2 new examples:
1)
Same process as before, using the same image and a grid ratio of
40:80, or 1:2. The few totally black areas are where no match of mean
averages could be found:
http://mwp.jaycloidt.com/MPIMAGES2006/W2x12006.jpg
2a)
Same process as before, using the same image and a grid ratio of
40:80, or 1:2, with a more diverse pool of source material:
http://mwp.jaycloidt.com/MPIMAGES2006/W2x22006.jpg
2b)
Detail from the second image above:
http://mwp.jaycloidt.com/MPIMAGES2006/W2x2DETAIL2006.jpg
mwp
On Sep 21, 2006, at 12:47 PM, mpalmer wrote:
The Newly-Created Woman 01
2006
An image from De Kooning’s classic Woman series is selected as
source material. It is then broken into grid-segments, in this case
80. The mean average of each segment is then matched with the
original image at a differently segmented grid-value, in this case
50. The match is carried out at three levels of identity:
1) Hue, Saturation and Value (Intensity).
If no match, then
2) Hue, Saturation.
If no match, then
3) Hue.
If still no match, the segment is left blank. In this particular
case, all segments are filled.
(Presentation format is the original on the left and the
transformed image on the right.)
http://mwp.jaycloidt.com/MPIMAGES2006/W12006.jpg
(Approx. 1MB)
mwp
