Real time implementation of an MRF-based motion detection algorithmon a DSP board

摘要

The main concern in image processing is computation cost. Markovrandom fields (MRF) based algorithms particularly require a significantcomputation cost. Most of implementations of this kind of algorithms aremade on parallel machines. This paper investigates an original solutionfor real time implementation of a robust MRF-based motion detectionalgorithm. A PC board, based on a pipeline architecture using a singlepowerfull DSP and FPGA components, is developed. The algorithm and theboard are described. A processing rate of 15 images per second isachieved, showing the validity of this approach