Hierarchical foveal algorithm development for ATR

Foveal active vision features imaging sensors and signal processing with graded acuity coupled with context sensitive gaze control, analogous to that prevalent throughout vertebrate vision. Foveal vision operates more efficiently in dynamic scenarios thatn uniform acuity vision because resolution is treated as a dynamically allocatable resource. Wide FOV and localized (central) high acquity are simultaneously supported while minimizing data to that which is relevant to the task. Efforts investigating machine implementations of foveal vision typically use polar sampling, which hampers signal processing, gaze control, and overall system development by yielding non-hierarchical data structures which are difficult to build and process. The proposed program will develop a computational Hierarchical Foveal Machine Vision (HFMV) sensor for Automatic Target Recognition (ATR) with monolithic spactiotemporal filtering, electronic gazing, and semi-parallel readouts. The sensor topology will support existing HFMV techniques, including multiacuity vision algorithms and gaze control, and many conventional hierarchical algorithms and processing hardware. The sensor topology will be optimized for ATR: the wide FOV and fast frame rate will support detection and tracking, and the high resolution foveal will demonstrate recognition.