No abstract

No abstract

A new fast block-matching algorithm for motion estimation is introduced. This algorithm uses a new enhanced motion vector system combmed with an effective M3 matching technique. The proposed method transforms the expensive 2-D block-matching problem into a simpler I-D matching by choosing some block as eligible candidates and eliminating the others. Test results of applying the proposed method on a number of MF' EG video sequences are included. These results indicate that the proposed method not only can reduce the computational complexity by a factor ranging h m 1 to 7, but also can reduce the prediction ermr by 2 to 20 percent as compared to the full search method. There are smne situations in which existing motion estimation tecbniques obtain a false motion vector with a huge prediction enor, while the proposed method can find an enhanced motion vector with an excellent accuracy.

This paper presents a novel bi-directional motion estimation technique, which determines the motion vectors more accurately in situations when there are variations in the inter-frame characteristics due to the changes in the brightness values of the corresponding pixels from frame to frame. The proposed bi-directional motion estimation technique is applied specifically to the fixed-size block, variable-size block and region-wise motion compensation schemes. The proposed method is applied to a number of benchmark video sequences and the results are compared with those obtained by applying the existing methods. These results show that the proposed method can improve the rate-distortion performance. In particular, in situations when there are large variations in the brightness of the corresponding regions from frame to frame, the proposed method improves significantly the rate-distortion performance.

In this paper, a new technique for reducing the blocking effect has .been presented. This technique cant be applied to different types of partitioning schenmes. In the proposed technique, the weighting coefficients of the OBMC scheme are modeled with four fuzzy relations. These fuzzy relations are formed based on two fuzzy sets, which define the closeness of a pixel to the block boundaries in both the horizontal and vertical directions. Furthermore, the appropriate fuzzy operators for these weighting functions are found. Simulation conducted on several test video sequences has demonstrated that the proposed technique can reduce the prediction error and improve the suibjective visual quality of a frame in comparison to the results obtained by using the generalized overlapped motion compensation scheme.