Hello, /sci/, programmer here looking for help solving a math equation so I can implement it into one of my apps.
The problem: You have an image of width W and height H (WxH) and you'd like to make as many largest size square crops of it possible. By necessity, since images on a computer can only be rectangles, at least, to my knowledge, this means a number of crops of square that are equal in side length to the largest side.
So, if you have a 512x256 rectangle, you will create a number of 256x256 squares separated by some distance of pixels each (uhhh Z). The end user can input the amount of pixels - smaller amounts will produce more crops for a larger fidelity end image but also be obviously more resource intensive since more images are created and used.
so, for Z = 256, you create a crop at (0, 256), and then a crop at (256, 256), for a total of 2 images. N = 2
Z = 128, you create a crop at (0, 256), a crop at (128, 256), and a crop at (256, 256). N = 3
Z = 64, you create a crop at (0, 256), a crop at (64, 256), etc., N = 5.
So as Z is halved (in this particular instance with a 2:1 aspect ratio image), N increases at a rate of [(2^N) + 1]. I can understand that well enough.
What I am having difficulty is generalizing this to something I can plug into anything in order to make it parameterizable and thus user configurable. Any assistance would be appreciated.
MS paint example for you if it helps you visualize it.
The problem: You have an image of width W and height H (WxH) and you'd like to make as many largest size square crops of it possible. By necessity, since images on a computer can only be rectangles, at least, to my knowledge, this means a number of crops of square that are equal in side length to the largest side.
So, if you have a 512x256 rectangle, you will create a number of 256x256 squares separated by some distance of pixels each (uhhh Z). The end user can input the amount of pixels - smaller amounts will produce more crops for a larger fidelity end image but also be obviously more resource intensive since more images are created and used.
so, for Z = 256, you create a crop at (0, 256), and then a crop at (256, 256), for a total of 2 images. N = 2
Z = 128, you create a crop at (0, 256), a crop at (128, 256), and a crop at (256, 256). N = 3
Z = 64, you create a crop at (0, 256), a crop at (64, 256), etc., N = 5.
So as Z is halved (in this particular instance with a 2:1 aspect ratio image), N increases at a rate of [(2^N) + 1]. I can understand that well enough.
What I am having difficulty is generalizing this to something I can plug into anything in order to make it parameterizable and thus user configurable. Any assistance would be appreciated.
MS paint example for you if it helps you visualize it.
