A Survey of Optimization of Rate-Distortion Techniques for H.264 Scalable Video Coding

Abstract

Rate-Distortion optimization is a process of improving a video quality during video compression. This will mainly concentrate on amount called rate and is a measure of distortion against data required to encode the video. The main usage of video encoders is to improve quality in encoding image, video, audio, and others to decrease the file size and to improve the quality. The classical method of making encoding decisions for the video encoder is to choose the result that yields highest quality output. Scalable video coding (SVC) is a process in which encoding and decoding can be done on a bit stream. The resulting video has lower temporal or spatial resolution or a reduced fidelity, while retaining reconstruction quality that is close to the original. SVC reduces bit rate necessarily required to represent a given level of perceptual quality and also improves coding efficiency. The main advantage of this Rate-Distortion is applied to measure the quality obtained from the video signals. Some of the rate-distortion models are prediction, approximation, Lagrangian multiplier, empirical, and parametric. Among these Lagrangian multiplier provides good rate-distortion optimization. The MSE- or PSNR-based quality metrics are usually used to assess the visual quality in the cases that the spatial resolution and frame rate are fixed. Our main aim is to demonstrate gain in PSNR with optimized value of bit rates. Finally, we conclude by presenting the efficient techniques to achieve improved quality of video.