Verification of Motion Estimator using Universal Verification Methodology (UVM)

Motion Estimation is a process to determine the motion vectors that describe transformation one 2D image to another; usually from adjacent frames in the video sequence. To obtain motion estimation block matching algorithm is used. Usually, the only difference between much of content in subsequent frames is motion. The task is to detect blocks of video data from successive frames that are related only via translation. This project aims to verify the design to perform the full search of Motion Estimator using Universal Verification Methodology (UVM). UVM is based on library classes and is an open source verification methodology. It is used for functional verification of digital hardware, primarily using simulation. UVM uses an architecture to build the verification environment that enhances reuse-ability. The main purpose of this project is to 100% coverage.