In english: you are breaking your image into 2 by 2 pixel regions and assigning each pixel in a region a value 0 1 2 3 (upper left, upper right, lower left, lower right) and then extracting that pixel from each of the regions, for your result.
In your current implementation, you pick a random value for the pixel choice, and then you use it on the entire region. But that value is not a parameter for the verb 'sample'. Instead you give a value which the choice must be less than, and you have sample pick an arbitrary value from the resulting range. Why? -- Raul 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=) > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
