Previous | Next --- Slide 7 of 36
Back to Lecture Thumbnails

Isn't familiarity the main reason we like regular grids? I think other grids can have all these properties, but we are simply not used to working with them.


Is ease of learning also related to this idea of simplicity? Or is it just easy to learn because we're familiar, because it's become entrenched?


I think hexagons probably work well too? But squares are just easier to think about or (maybe) to manufacture.


another good reason is probably for convolution in image processing, if they are squares it's easy to know the 8 neighbors of a square pixel


Isn't hexagons also generalizable? I think we can encode basically any image given enough pixels of any shape i.e. by increasing resolution.


Why didn't we use a regular grid of 4 triangles composing our typical pixels?


What is anisotropy and why regular grids cause it?