INTRODUCTION To achieve very low bit-rate transmission, many video coding
Video coding has gained increasing interest in the past papers suggest extraction of stationary objects and the backfew years. The temporal decimation technique is regarded ground and coding only of moving objects. These techniques, which could seriously distort the output, are mostly applied to as a powerful tool for increasing compression for video spatial domain. A novel scalable adaptive temporal segmentacoding. Spatial decimation algorithms can also be used in tion algorithm for video coding is proposed in this paper. Our conjunction with other video processing techniques such proposed algorithm is scalable in terms of compression, quality, as using the motion vector estimation technique [1, 2] for and some specific features. A multifunction control of the outa further improvement of the compression ratio. The temput is formulated as a cost function, which is flexible enough poral decimation approach became mature recently when to implement any desirable criteria, such as channel capacity or the advancement of this approach using Olstad's adaptive reconstruction quality. Our algorithm uses a simple but efficient temporal decimation [3] made some significant contribuquantization process instead of using a computationally expentions. The approach potentially could achieve very low bitsive discontinuity detection. The approach can also be regarded rate video coding.
as a combination of adaptive temporal decimation and seg-
Obviously, a simple temporal decimation can be mented video coding. The overall bit rate can be regulated according to the user's specifications. Experimental results achieved by periodically dropping whole frames from imshow that the overall bit rate of video coding can be controlled age sequences. However, this may cause objectionable easily using our scheme. The output quality is good for both temporal aliasing artifacts, especially in regions where temsubjective and objective tests. We have also found that the size poral activity is high. Olstad's algorithm adaptively selects of the time window (N) has significant effects on the quality important blocks for decimation, resulting in reconstructed of the reconstruction, while our proposed algorithm outpersequences of better quality. forms the adaptive temporal decimation algorithm. ยฉ 1997 Aca-Although temporal decimation algorithms can achieve demic Press * To whom correspondence should be addressed at the Department blocks; (3) portability; (4) ineffective coding of the Block of Electronic Engineering, the Hong Kong Polytechnic University, Hung Position Map (BPM); and (5) too many Intra frames.