To get a better idea of the real "specimen", consider the following noun, ss where you can begin to see a capital J. To recreate a specimen in order to exercise the earlier posted compressions, consider using the code which follows ss and converts ss into specimen.
ss =: 0 : 0 000000000000 000000000000 000000000000 000000000000 000000000000 000000000000 00007FFF8000 00007FFF8000 00007FFF8000 00007FFF8000 00007FFF0000 000003E00000 000003E00000 000003E00000 000003E00000 000003E00000 000003E00000 000003E00000 000003E00000 000001E00000 000001E00000 000001E00000 000001E00000 000001E00000 000001E00000 000001E00000 000001E00000 000003E00000 000003C00000 1F0003C00000 1E0003C00000 1E0003C00000 1E0003C00000 1E0003C00000 1E0003C00000 1F0003C00000 0F0007C00000 0F000F800000 0FC01F800000 07FFFF000000 07FFFF000000 03FFFE000000 00FFF8000000 000000000000 000000000000 000000000000 000000000000 000000000000 000000000000 000000000000 000000000000 000000000000 ) dfh=: 16 #. 16 | '0123456789ABCDEF0123456789abcdef' i. ] specimen=: (#: dfh;._2 ss),"1] 7#0 On Mon, Dec 31, 2012 at 6:07 PM, Brian Schott <[email protected]> wrote: > I developed a small image compression script which is applied to a > "thresholded" image from the G component only of the RGB triplet. (Although > my example below is of an 8x8 thresholded image, in practice it is 52x52.) > > wind=: 2 2$2 > sample=: 4 : 'wind (?x)&{@,;._3"2 y' > ]specimen=: 8 8?.@$2 NB. the test thresholded image > 0 1 0 1 1 1 0 0 > 0 1 0 1 0 0 1 0 > 0 0 0 0 1 1 1 0 > 0 0 0 0 1 1 0 1 > 1 0 0 1 0 0 1 0 > 1 0 0 0 0 1 0 1 > 0 0 1 0 0 1 0 0 > 1 0 0 1 0 1 0 0 > 2 (4&sample) specimen NB. a random resulting compression > > Eventually I will have to do this compression in C and I am very > weak in C, so I am looking for an alternative calculation that gives > approximately the same result but using something like matrix > multiplication. I am just beginning to learn the terminology of image > transformation, but this may be an example of a "convolution" in that > discipline. > > So, my question is (I think) can a random convolution matrix be > constructed to accomplish a similar result using matrix multiplication, for > which I have found a c code example at rosettacode.org? > > http://rosettacode.org/wiki/Matrix_multiplication#C > > Thanks, > > (B=) -- (B=) ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
