Skip to Main Content
In conventional video coding, the complexity of an encoder is generally much higher than that of a decoder because operations, such as motion estimation, consume significant computational resources. Such a codec architecture is suitable for a broadcast downlink transmission model. However, in contemporary applications of mobile wireless video uplink transmission, it is desirable to have a low complexity video encoder to meet the resource limitations on mobile devices. Recent advances in distributed video source coding provide a potential reverse in computational complexity for encoder and decoder (Puri, R. and Ramchandran, K., 2002; Aaron, A. and Girod, B., 2003). Various studies are done within the context of a scheme that is similar in spirit to that of Aaron and Girod. We conduct a study of encoding and decoding techniques for syndrome-based video encoding schemes for mobile wireless applications. Our innovation is the adoption of low resolution low quality reference frames for motion estimation at the decoder. To accomplish the goal of reducing computational complexity while maintaining good reconstruction quality at the decoder, we investigate the following related strategies: (1) compression of low resolution and low quality sequences; (2) rate-distortion study with complexity constraints at the encoder; (3) building enhanced evidence for a syndrome decoder. Extensive experimental results have confirmed the effectiveness of these techniques for syndrome based video coding.