Proposed One Point Pentagon inner search Fast Block Matching Algorithm
Imbaby Ismail Mahmoud, Sherief Hashima, Atef Abou Elazm · National Radio Science Conference · 2009
Fast Block Matching Algorithms (BMAs) play an important role in motion estimation and object tracking. Search speed and distortion performance are determined by search pattern shape and size. A fast motion estimation algorithm, called One Point Pentagon inner search (OPPEN) is presented for both motion estimation and object tracking. The performance of the OPPEN algorithm is analyzed and quantitatively compared with most recent BMAs. Experimental results reveal that our proposed algorithm has superior performance compared with other algorithms in terms of number of search points and computational complexity. Also, the role of Region of Interest (ROI) in object tracking is discussed for further improving the performance of the proposed algorithm.