Parallel issues of an optical flow estimator based on a quasi-interpolant operator

This paper addresses the parallel issues of a novel method entirely developed by the authors devoted to estimate the velocity vector field of the apparent motion of brightness patterns usually referred as optical flow. Optical flow plays a fundamental role in many real-time computer vision applications. In this paper the flow field has been performed by adopting a bivariate quasi-interpolant operator based on centered cardinal B-spline functions onto a non linear minimizing algorithm. The solving model involves massive computational amount so that high-performance system need to rapidly proceed in the computations. Therefore, in the following a parallel analysis of the computational scheme, oriented to distributed multiprocessors platforms, has been carried out. The process has shown to be synchronous, with good tasks balancing and requiring few amount of data transfer.