Image Processing Algorithms for Target Discrimination

MDA seeks for an efficient target tracking and discrimination system beyond the current state of the art. The primary motivation of the work proposed is to offer a topographic microprocessor architecture for multi-target discrimination and tracking with embedded sensors capable of operating in a process real-time manner. Phase I effort has shown that the performance of multi-target tracking (MTT) and discrimination (MTD) systems can be significantly increased with stored program adaptive cellular array sensors. In these experiments the input data flow acquired by the array sensors is processed on an adaptive cellular visual microprocessor (CVM) architecture consisting of both cellular nonlinear network (CNN) and digital signal processing (DSP) microprocessors. The algorithms designed for this combined hardware platform use adaptive multi-channel CNN solutions for instantaneous position estimation and morphological characterization of all targets; and DSP solutions for distance calculation, gating, data association, track maintenance and dynamic target motion prediction. A special feature of the architecture is that it allows an interactive communication between the sensor-level and the digital environment. The system development includes both the chip design (a scalable and locally adaptive CVM/CNN sensor-processor array) and the associated computational infrastructure (a compact standalone cellular vision system).