Applications of systolic array architectures to motion analysis studies

Several types of systolic array architecture with applications in motion analysis studies are presented. These architectures are designed for rapid computation of velocity, acceleration, and jerk. Once the pipeline is full, the values of these parameters are obtained at every clock cycle. For the linear array, (n+1) processing cells are required for computing a derivative of order (n). Nine processing cells are needed for the triangular systolic array for computing position velocity, acceleration, and jerk.<>