3/22 Button Function, Images and Tile Details, and Algorithm Thinking

Done:

I made a function that is called when each button is clicked. This function slices the respective image and prints some relevant points to the console regarding that image.

I also met with Dr. McVey. We talked about scaling, re-sizing, image sizes, and tile numbers. She advised that consistent scaling will work fine, that it is fine to force a certain image size for now (through scaling or re-sizing), and that tile amounts may be static or dependent on user input.

We also discussed that a recursive algorithm that varies the tile subdivisions depending on color density might not be necessary. Not only is it complex, but there is a way to improve upon a fitness function that merely compares the average color of tiles. If I do pixel-by-pixel, or comparing a handful of pixels at different spots, or even the average of some pixels in a few areas (e.g., top-left, top-right), these can make more accurate comparisons. With any of these, I will be more able to catch differences around the tile, if any, to varying degrees of accuracy vs. speed.

Next Steps

• Some computations

• Generate member of the population

• How to store them