Motion Estimation for Particle Images in a Finite Rate of Innovation Framework

Abstract:
In fluid mechanics a major field is velocity measurement based on particle motion in image sequences. These techniques suffer from the image acquisition with high temporal and spatial resolution that restrict the possible observation time due to memory constraints. This paper proposes a sparse signal processing approach for the imaging process. A specific system design based on Finite Rate of Innovation Signals is proposed. Performance bounds for the motion estimation of the particles are given in terms of the Cram´er-Rao bound. Furthermore, two algorithms are proposed to estimate the particle motion. Numerical results are provided to compare the performance of the algorithms with theoretical bounds.