More recently, Ferreira et al. [42] proposed a shot-based key-frame extraction method based on rate-distortion optimisation for 2D and 3D video.
Based on rate-distortion results, the conventional decimation method and the high-performance decimation method gave the same coding efficiency.
Based on the accuracies of the SI frames, the best coding mode is selected based on rate-distortion (RD) optimization; thus, the GOP size can be adaptively and hierarchically set.
Huang et al. [51] also presented a key-frame extraction method for 3D video based on rate-distortion optimisation, where Rate and Distortion definitions are similar to those used in [19].
However, to achieve this, the H.264 encoder employs a complex mode decision technique based on rate-distortion optimization (RDO) that requires high computational complexity, which significantly increases the encoder complexity.
In view of this drawback, we propose here a unified rate allocation scheme, based on rate-distortion-smoothness optimization criterion and multi-stage dynamic programming (DP) technique, to solve the above problems.
Then, These methods based on rate (congestion -distortion optimized pacongestion -distortion theoptimizedy to packete an optimal schedulingo mediareware scheduling problem.
The quantization distortion of the receive node i depends on Q, and based on the rate-distortion theory [14] is equal to D i = ( P i rec + σ 2 ) 2 − Q, (16).
Based on power-rate-distortion (P-R-D) model in [10], they prove that power is tightly coupled with rate, thus, to trade bits for joules and to perform energy minimization are rapid method to obtain minimum energy [11].
For residual coding, the transform block sizes and shapes are determined based on a rate-distortion optimization (RDO).
