@alisavv For every pixel in the image, take the difference in red, green, and blue values, square them, and then sum them over every pixel. So if 0 is the original image and 1 is the new image, Sum [(R1-R0)^2 + (G1-G0)^2 + (B1-B0)^2] for ever pixel in the image. Then, the lower the sum, the more alike the two images are.
This is pretty cool. If you slowed it down and narrated the theory behind it this could be a perfect intro to genetic algorithms for the total beginner.
This video is missing the first 1000 generations which is where most interesting things happen. I will post a new video over the weekend showing what happens during the first few 1000 iterations.
how to comparing image's similarity?
alisavv 11 months ago
@alisavv For every pixel in the image, take the difference in red, green, and blue values, square them, and then sum them over every pixel. So if 0 is the original image and 1 is the new image, Sum [(R1-R0)^2 + (G1-G0)^2 + (B1-B0)^2] for ever pixel in the image. Then, the lower the sum, the more alike the two images are.
thomasageorgiou 11 months ago 2
@thomasageorgiou thanks. very cleverly solution :)
alisavv 11 months ago
nice video
cracylord 1 year ago
This is pretty cool. If you slowed it down and narrated the theory behind it this could be a perfect intro to genetic algorithms for the total beginner.
kdejeff 3 years ago
@kdejeff Please no.
TimJSwan89 8 months ago
This video is missing the first 1000 generations which is where most interesting things happen. I will post a new video over the weekend showing what happens during the first few 1000 iterations.
thomasageorgiou 3 years ago
That's sweet :D
WhiteDragon103 3 years ago