Boundary-Control Vector (BCV) Motion Field Representation and Estimation by Using a Markov Random Field Model1

DFD have to be transmitted to the receiver, a well designed video coder should balance the bits used in these two A new motion field representation based on the boundarycontrol vector (BCV) scheme for video coding is examined in parts. Other factors of consideration in video coder design this work. With this scheme, the motion field is characterized include computational cost, hardware complexity and the by a set of control vectors and boundary functions. The control domain of applicability. vectors are associated with the center points of blocks to control We can roughly classify existing motion field representathe overall motion behavior. We use the boundary functions tion into block-based, pel-based, and model-based categoto specify the continuity of the motion field across adjacent ries. The block-based representation has been widely used blocks. For BCV-based motion field estimation, an optimization and adopted by several standards such as H.261 [16] and framework based on the Markov random field model and maxiMPEG [6]. It divides an image frame into nonoverlapping mum a posteriori (MAP) criterion is used. The new scheme blocks, and represents the motion field in each block with effectively represents complex motions such as translation, roa translation vector. This representation is generally applitation, zooming, and deformation and does not require complex scene analysis. Compared with MPEG of similar decoded SNR cable and concise. A differential coding can be used to (signal-to-noise ratio) quality, 15–65% bit rate saving can be further reduce the redundancy between motion vectors achieved in the proposed scheme with a more pleasant visual by exploiting their spatial correlation. The block-based quality.  1996 Academic Press, Inc. motion field can be estimated by using a straightforward block matching algorithm (BMA) or its variants. However, the block-based scheme has some limitations. The tradi