Research And Implementation Of Scalable Video Coding Technology

The scalable video coding was advanced in the MPEG-2, the basic idea of which is the layer coding for the one frame. Scalable video coding includes the base layer which contains the basic information of the frame and the enhancement layer which contains the remainder information of the frame.The base layer and the enhancement layer are relative with each other. The base layer is decoded by itself while the enhancement layer is decoded by itself and the base layer information. In the low bandwidth environment, the encoder can transmit the base layer only, while in the high bandwidth environment, the encoder can transmit both the base layer and the enhancement layers.There are five basic methods of sclable video coding such as the spatial scalability,temporal scalbility,SNR,data partition and the fine grain scalable (FGS). Besides the scalabilities mentioned as above, there are the scalability based on the shape coding and the scalability based on the H.264/AVC. In the practical application, the solution of the scalable video coding may be the combination of two or more methods mentioned as above.Traditional scalable video encoders sacrifice coding efficiency to reduce error propagation because they have avoided using high quality reference layer information to predict the low quality reference information to prevent the error propagation termed "drift". In this paper, we propose an optimized H.264/AVC based fine grain scalable (FGS) video encoder which achieves the significant better compression efficiency than the ordinary FGS encoder. It intentionally allows the drift produced by predicting the low quality layer from the high quality layer. Our drift control system balances the tradeoff between compression efficiency and error propagation. The proposed scalable coder uses a spatially adaptive procedure that optimally selects key encoder parameters: the quantization, the prediction strategy, packet loss, error concealment. Also.we propose a method to measure the drift based on the statistics by calculate the expectation of the differences between original and the reconstructed 4×4 macroblocks.