Data association techniques for advanced driver assistance systems using embedded soft-core processors

Advanced Driver Assistance Systems (ADAS) for automotive applications are used to reduce the number of road accidents. Multiple Target Tracking (MTT) is one of the most efficient techniques used in modern ADAS's. Data Association is a vital part of MTT which can be modeled as the classic Assignment Problem. We implement two algorithms namely, the Munkres (or Hungarian) algorithm and the Auction algorithm in Altera's StratixII FPGA using the NiosII soft-core processor. We analyze the resource requirements and the execution speed of the two algorithms. We describe the circumstances where the use of one or the other of these two algorithms is most suitable in the context of the data association problem.