[in Japanese]

View full abstract

The method proposed in this paper uses the area of the region around each pixel with small matching errors as a criterion for motion estimation and segmentation. This area is calculated for each displacement vector through directional distance transformation of a binary image obtained by threshoulding the matching error at each pixel. On the basis of this area, the proposed method selects candidate vectors at each pixel and determines one of them to be the motion vector by majority rule in a local region. As a result, an image is divided into regions with uniform motion. This method has the advantage of being insensitive to surface textures since it uses only matching errors. The simulation results show that this method can perform well both motion estimation and segmentation even along motion boundaries when regions with coherent motion are relatively large.

View full abstract

Currently, various kinds of very low bit-rate coding schemes are being studied. One of the schemes is MC-DCT interframe coding. In this scheme, only the data of significant blocks is transmitted. So far, the block significance designation criteria have been quite simple. For example, the decision may depend on a function evaluating the interframe difference signal. When the noise signal is large, noisy background blocks with large interframe difference signals are designated significant at the expense of blocks on, for example human subject, with small interframe difference signals.
Here we propose a criterion, in which blocks on the subject region are preferentially designated significant by fuzzy clustering and fuzzy inferences. We start with the subject being located in the center of the first frame and then track the subject's movement through subsequent frames. Simulation experiments show that the proposed method achieves good results.

View full abstract

In this paper, we will examine, through subjective assessment tests, the effectiveness of the bi-directional prediction scheme, which was adopted by MPEG as a standardized coding technique and which is generally assumed to be effective with respect to the coding efficiency. We will also clarify the distinctive picture quality degradation by way of the bi-directional prediction coding scheme, analyze causes, and evaluate the de7gradation quantitatively. Specifically, the assessment results show that the bi-directional prediction scheme works well in most of the sequences with respect to the picture quality improvement, but causes flickers to some sequences. The analyses show that for the picture with high contrast texture the resolution of the decoded image differs between the B-pictures and the P-pictures; this leads to the flicker through the gamma characteristics of a monitor. We will demonstrate that we can evaluate the flicker by simulating the electro-optic conversion process.

View full abstract