Pyramid vision applied to automatic inspection

A processing structure that is variously called a pyramid, a quadtree, a recognition cone, or a processing cone has been proposed as a means of achieving a high degree of parallelism for tasks such as object recognition in machine vision. In its simplest form the pyramid consists of a base with p^m by pⁿ pixels, where typically p=2 or p=3. This is called a `retinal memory´