Maximum detection with a two-dimensional optoelectronic p-n-p-n winner-take-all network.

An optoelectronic p-n-p-n winner-take-all network is presented that can identify the location of the maximum intensity in two-dimensional spatially distributed light. An input pattern of illumination is incident upon an array of p-n-p-n devices that are connected in parallel to a common load resistor and therefore have identical voltage biases. If the light intensity at one or several elements exceeds a threshold value, only one p-n-p-n element stays on and emits light when the input light pattern is turned off. The element that is turned on is the one that receives the maximum illumination.