Abstract: A state based control system (100) is provided for use with unreliable physical device sensor inputs. The system (100) includes symmetrical architectures between a client side and a server side. Communications between both architectures are state dependent thus minimizing conflicts, and reducing system reliance on consistent device signaling. Physical devices (144, 146) are represented by surrogates which increase system reliability, deepen complexity, and simplify end user experience.

Abstract: A method of shielding light from rows and columns of an array of pixels of a CMOS image sensor includes the steps of: providing a first layer of metal shielding above the rows and columns of the array; and providing a second layer of metal shielding above the first layer. The method further includes forming slots within each of the first and second layers of metal shielding, such that the slots of the first layer are misaligned with the slots of the second layer. A light shield for shielding columns and rows of an array of pixels in a CMOS image sensor includes a first metal layer located above the columns and rows of the array, and a second metal layer located above the first metal layer. Each of the first and second metal layers includes slots. The slots are formed such that the slots of the first layer are misaligned with the slots of the second layer.

Abstract: A simple transistor circuit which acts as a linear resistor for small applied voltages, but becomes extremely resistive for large applied voltages is disclosed. Two-dimensional resistive grids comprising these resistive fuses can be employed to smooth and segment discretized images in machine vision. Existing and previously proposed VLSI implementations of resistive fuses have required at least thirty-three transistors. The resistive fuse circuit of the present invention uses only four transistors in its simplest embodiment, thus making it possible to design much denser vision arrays.