[Paper Review] Optimization of the Block-level Bit Allocation in Perceptual Video Coding based on MINMAX

Chao Wang, Xuanqin Mou+1 more|arXiv (Cornell University)|Nov 15, 2015

Video Coding and Compression Technologies38 references3 citations

TL;DR

This paper proposes a macroblock-level bit allocation method for perceptual video coding using a MINMAX criterion to minimize quality fluctuation across blocks. By iteratively optimizing quantization parameters via a Maximum Distortion Descend (MDD) algorithm and using the SSIM index to measure perceptual distortion, the method achieves significant bit savings and improved visual quality on benchmark sequences with acceptable computational complexity.

ABSTRACT

In video coding, it is expected that the encoder could adaptively select the encoding parameters (e.g., quantization parameter) to optimize the bit allocation to different sources under the given constraint. However, in hybrid video coding, the dependency between sources brings high complexity for the bit allocation optimization, especially in the block-level, and existing optimization methods mostly focus on frame-level bit allocation. In this paper, we propose a macroblock (MB) level bit allocation method based on the minimum maximum (MINMAX) criterion, which has acceptable encoding complexity for offline applications. An iterative-based algorithm, namely maximum distortion descend (MDD), is developed to reduce quality fluctuation among MBs within a frame, where the Structure SIMilarity (SSIM) index is used to measure the perceptual distortion of MBs. Our extensive experimental results on benchmark video sequences show that the proposed method can greatly enhance the encoding performance in terms of both bits saving and perceptual quality improvement.

Motivation & Objective

• To address the challenge of high complexity in block-level bit allocation due to interdependencies between video coding units.
• To reduce perceptual quality fluctuation among macroblocks within a frame during video encoding.
• To develop a low-complexity, offline-optimized bit allocation strategy suitable for perceptual video coding.
• To improve overall encoding efficiency by balancing bit rate and perceptual quality at the macroblock level.
• To extend existing frame-level bit allocation methods to the more granular macroblock level with perceptual fidelity.

Proposed method

• The method employs a MINMAX criterion to minimize the maximum distortion across all macroblocks in a frame.
• An iterative algorithm called Maximum Distortion Descend (MDD) is used to adjust quantization parameters across macroblocks to reduce peak distortion.
• The Structural Similarity (SSIM) index is adopted as the perceptual distortion metric to evaluate macroblock quality.
• Bit allocation is optimized at the macroblock level rather than at the frame level, allowing finer-grained control over quality distribution.
• The optimization process is designed to be computationally feasible for offline video encoding applications.
• The method dynamically redistributes bits across macroblocks to minimize the worst-case perceptual distortion while respecting total bit budget constraints.

Experimental results

Research questions

• RQ1 How can block-level bit allocation be optimized to minimize perceptual quality variation across macroblocks in video coding?
• RQ2 What is the impact of using the MINMAX criterion on perceptual video quality and bit rate efficiency at the macroblock level?
• RQ3 Can an iterative optimization algorithm reduce quality fluctuation while maintaining low computational complexity?
• RQ4 How does the use of SSIM as a perceptual distortion metric compare to traditional metrics in block-level bit allocation?
• RQ5 To what extent can bit savings be achieved without sacrificing perceptual quality using this macroblock-level approach?

Key findings

• The proposed method significantly reduces quality fluctuation among macroblocks by minimizing the maximum SSIM distortion across the frame.
• Extensive experiments on benchmark video sequences demonstrate substantial bit savings while maintaining or improving perceptual quality.
• The method achieves better perceptual quality compared to conventional frame-level bit allocation techniques, as validated by SSIM and subjective evaluation.
• The encoding complexity remains acceptable for offline applications, making the approach practical for high-quality video encoding.
• The MDD algorithm effectively converges to a near-optimal bit allocation solution with minimal distortion variation.
• The results show that macroblock-level optimization leads to more consistent visual quality across the entire video frame.