Abstract

Imagery edges occur naturally in human visual systems as a consequence of redundancy reduction towards “sparse & orthogonality feature maps,” which have been recently derived from the maximum entropy information-theoretical first principle of artificial neural networks. After a brief math review of such an Independent Component Analysis (ICA) or Blind Source Separation (BSS) of edge maps, we explore the de-mixing condition for more than two imagery objects recognizable by an intelligent pair of cameras with memory in a time-multiplex fashion..

We assume that during the first look the system has enough time to have two cameras pointed, triangulated, for instance, focused on the girl_G standing in the shadow of a tree_T hidden with birds_B. Mathematically these objects are mixed in two cameras with 2x2-matrix [A₀] such as: (x₁ x₂)^T=[A₀](G , t+εb)^t without knowing birds hidden quietly in the tree indicated by a small order of magnitude 0(ε). Suddenly the birds begin to sing. By design, two-camera system has two separate pointing & focusing & memory capabilities. Since as a quick response, only one of the dominate camera having the de-mixed girl image G can point away and focus at the birds hidden in the tree with a second look defined mathematically by (y₁ blurred y₂)^T = [A₁](T, b)^t. By design, the new image y₁ of birds & tree is acquired by the dominating camera with a new focus, whereas the other has a blurred y₂ and is therefore replenished by the previous de-mixed image (T+εB), provided that each camera processes its own memory update rule. We consider channel communication application that we can efficiently mix four images using matrices [A₀] and [A₁] to send through two channels.

© 2000 SPIE