ADVANCED SYSTEMS AND TECHNOLOGIES FOR FUTURE NAVAL WARFARE HIERARCHICAL FOVEAL MACHINE VISION SYSTEMS DEVELOPMENT FOR ATR

A fundamental problem in automatic target recognition (ATR) is the overwhelming amount of sensor data which must be processed. Target features are localized within the field-of-view (FOV) of the imaging sensor. Uniformly sampling within the FOV is thuus inappropriate, regions with little or no relevance to the task are samples at the same resolution as key features, occupying valuable signal bandwidth and computational resources, and increasing system latencies. Foveal active vision features imaging sensors and processing with graded acuity coupled with context sensitive gaze control, analogous to that prvalent throughout vertebrate vision. Foveal vision operates more efficiently in dynamic scenarios than uniform acuity vision because resolution is treated as a dynmaically allocatable resource, simultaneously achieving wider field-of-view, greater localized resolution, and faster frame rates with significantly reduced processing requirements. The proposed program will develop commercially feasible foveal vision technology and integrate it into a Hierarchical Foveal Machine Vision System for ATR applications. The system uses rectilinear multiresolution sampling and processing which together exploit existing algorithms and hardware, ensure component feasiblity, accelerate prototyping and commercilaization, and reduce manufacturing costs. Hierarchical processing further reduces hardware requirements and accelerates system response.