mat_kinotek

I'm messing around with jit.boids and cv.jit.canny. My goal is to use the pixel values from cv.jit.snake as x,y coordinate to mix (with jit.xfade) with the output of jit.boids2D.

I realize that jit.boids2D output a 2 planes 500x1 matrix, so I will need to unroll the x values of the cv.jit.canny as plane 1 and y values as plane 2 of a new 2 planes matrix to mix with the jit.boids2D output.

I'm using jit.iter but I'm not able to figure out how exactly have the x and y coordinates in the 2 separate planes.

Anyone have tried this before? Any help is reallt apreciated!!

hello folks,
I'm messing around with jit.boids and cv.jit.canny. My goal is to use the pixel values from cv.jit.snake as x,y coordinate to mix (with jit.xfade) with the output of jit.boids2D.
I realize that jit.boids2D output a 2 planes 500x1 matrix, so I will need to unroll the x values of the cv.jit.canny as plane 1 and y values as plane 2 of a new 2 planes matrix to mix with the jit.boids2D output.
I'm using jit.iter but I'm not able to figure out how exactly have the x and y coordinates in the 2 separate planes.
Anyone have tried this before? Any help is reallt apreciated!!
/matt


boids2d-and-cv-jit-canny

you will have to figure out a solution for either not having enough white pixels per boid, or too many:

something like this?
you will have to figure out a solution for either not having enough white pixels per boid, or too many:


but it seems that the matrix feeded by jit.canny is"upside down" and smaller. But I dont get why. This is the patch with 2 object and xfade:

wow thanks robtherich!
this really helped a lot!
but it seems that the matrix feeded by jit.canny is"upside down" and smaller. But I dont get why. This is the patch with 2 object and xfade:


because of the difference in origin between jitter matrices (top left corner) and opengl (center), and direction of increasing Y axis (down for matrices, up for opengl).

boids2D and cv.jit.canny ?