In the video applications using the international video coding standards, it is necessary to develop a successful rate (or quality) control algorithm and new service schemes to provide pleasant video quality for viewers. In this dissertation, we propose some efficient quality control schemes for video coding based on rate-distortion (R-D) sense.
First, we propose an R-D estimation method for MPEG-2 video. In general, video coding standards, like MPEG-2, control the output bit rate or picture quality by adjusting the quantization parameter (QP) level. Therefore, accurate estimation of the rate and the distortion corresponding to the applied QP is helpful to the efficient and precise control of video encoding. The proposed R-D estimation method offers a closed-form mathematical model that enables us to predict the bits and the distortion generated from an encoded frame at a given QP level, and vice versa. Our R-D estimation model is derived from the observation of the relationship between non-zero coefficients of quantized DCT coefficients and quantization characteristics. Our R-D estimation requires just one frame delay and low computational complexity because its major operation is only to obtain a histogram or weighted histogram of DCT coefficients from an input frame. Furthermore, its results are accurate enough to be applied to practical video coding. In our simulation, the estimation errors for the rate and the distortion are less than 2.5% and 1.5%, respectively. Therefore, the proposed R-D estimation model is appropriate for the applications requiring low delay and low complexity, such as real-time bi-directional video communications.
Second, we propose a forward rate control scheme using the proposed R-D estimation results. The proposed rate control scheme ensures that the video buffers do not underflow and overflow by satisfying the buffer constraint and additionally prevents that quality difference between consecutive frames exceeds a demanded certain level by adopting the distortion constraint. In addition, a consistent picture quality is maintained within a frame by using the same QP level for the frame, and error propagation which is caused by quality degradation of anchor frames is reduced by differentiating the control procedure for the anchor frames from that for the non-anchor frames. Simulation results show that our control scheme achieves 0.52-1.84 dB peak signal-to-noise ratio (PSNR) gain over MPEG-2 Test Model 5 (TM5) rate control and maintains very consistent quality in a frame as well as between frames.
Finally, we present a joint quality control scheme to be able to accurately control the relative picture quality among the video programs in terms of PSNR. A conventional constant bit rate (CBR) channel is now capable of delivering several digitally compressed video programs due to recent advances in video compression and digital transmission technology. In this environment, if the relative picture quality among the programs can be discriminated according to the importance of each program, more improved video services can be provided to the viewers. In this aspect, we propose a joint quality control scheme using the proposed R-D estimation results. Our joint quality control scheme allows variable bit rate (VBR) for each video program to maintain the pre-determined relative picture quality among the aggregated video programs while keeping a constant sum of the bit rates for all programs to be transmitted over a single CBR channel. This is achieved by simultaneous controlling the video encoders to generate VBR video streams at the central controller. We discuss three types of joint quality control systems depending on the encoder and decoder buffers configuration. Furthermore we also suggest their buffer regulation methods based on the analysis of the constraints imposed by sender/receiver buffer sizes and total transmission rate. Through various simulation results, it is found that our quality control systems guarantee that the video buffers do not overflow and underflow and the quality control errors do not exceed 0.1 dB.
