Nonlinear competitive dynamics of a large array of opto-electronic feedback circuit systems for image processing applications

Fast, compact, efficient systems for high-resolution image analysis remain a challenge for image processing. These systems must be able to process high- or super-resolution images in real-time and provide data that maintains the important information for further processing, analysis, or decision-making. The resulting images represented in a reduced format can then be transmitted under limited bandwidth conditions. This is desirable, for example, in the battlefield environment when data is obtained in real-time and must be transmitted to a command/control center.;Opto-electronic systems provide highly parallel computing schemes that can be used for real-time image processing. This research proposes two architectures for a large-array of opto-electronic feedback circuits that can be realized using cutting-edge technology to perform edge detection, segmentation, and target tracking in high- or super-resolution images.;The first system uses a pixelated large array of opto-electronic feedback circuit based on competitive or winner-take-all (WTA) dynamics. Edge detection and image segmentation are demonstrated through simulations using system parameters that allow multiple winners, where the winning pixels represent the object of interest within an image.;The second system is based on nonlinear dynamics and pattern formation of a large array of opto-electronic feedback circuit. An optical soliton pattern structure and the competitive or WTA dynamics of the system are used in the processing scheme to locate and track a critical object within an image. Edge detection is also performed using soliton representation of objects within an image.;This research also presents a new and unique method, to potentially represent image processing in the visual system. Simulations show soliton monopoles and dipoles "acting" in a manner similar to groups of neurons responding to stimulus applied to the visual field. The soliton response can be seen as mimicking the response of cells found in the visual cortex allowing object location, edge detection, segmentation, and tracking.