Adaptive Motion Vector Control for SDR HDR Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In high-definition video encoding, particularly for 4K and 8K resolutions, the motion vector restriction at slice boundaries can lead to image quality degradation due to the inability to select an optimum motion vector, especially in fast-moving scenes, and this issue is exacerbated by the need to consider SDR/HDR switching.
Innovation Solution
A video encoding method that dynamically adjusts the scalable encoding structure based on dynamic range-related information to select an appropriate motion vector interval, allowing for seamless switching between SDR and HDR, thereby preventing image quality degradation by optimizing motion vector selection near slice boundaries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If motion vector restriction is applied at slice boundaries, then processing load is reduced and encoding efficiency is improved, but image quality degrades in fast-moving scenes due to inability to select optimum motion vector
Solution Approach 1:
The patent applies dynamics by making the motion vector restriction adaptive rather than static. The restriction level changes dynamically based on scene characteristics: in fast-moving scenes, the restriction is relaxed to allow larger motion vectors, while in slow-moving scenes, the restriction is maintained to preserve encoding efficiency. This resolves the contradiction by allowing the system to optimize between image quality and encoding efficiency depending on real-time conditions.
Solution Approach 2:
The patent changes the motion vector restriction parameter based on scene characteristics. Specifically, the maximum motion vector value at slice boundaries is adjusted according to the temporal complexity or motion activity detected in the scene. This parameter adaptation allows the system to maintain image quality in fast-moving scenes while preserving encoding efficiency in slower scenes, directly resolving the technical contradiction.
2Productivity
If large M value (large reference picture interval) is used, then coding efficiency increases in motionless scenes, but motion vector value becomes large which may exceed slice boundary constraints
Solution Approach 1:
The patent makes the reference picture interval (M value) adaptive rather than fixed. The system dynamically adjusts M based on scene characteristics: in motionless or slow-moving scenes, larger M values are used to improve coding efficiency, while in fast-moving scenes, smaller M values are selected to keep motion vectors within acceptable ranges and avoid exceeding slice boundary constraints. This dynamic adjustment resolves the contradiction between coding efficiency and motion vector magnitude.
Solution Approach 2:
The patent changes the reference picture interval parameter adaptively based on temporal complexity analysis. When the scene is determined to be motionless or slow-moving, the system increases M to improve compression efficiency. When fast motion is detected, the system decreases M to maintain motion vector values within constraints. This parameter adaptation directly addresses the contradiction between coding efficiency and motion vector magnitude.
3Power
If screen 4-division encoding is used, then processing load in 8K is reduced, but local image quality degradation occurs at slice boundaries due to motion vector constraints
Solution Approach 1:
The patent applies local quality by implementing different motion vector restriction levels at different locations within the screen. Specifically, blocks near slice boundaries are subject to adaptive motion vector restrictions based on local scene characteristics, while blocks away from boundaries follow standard encoding. This localized adaptation allows the system to maintain image quality at critical boundary regions while preserving the processing load reduction benefits of screen division encoding.
Solution Approach 2:
The patent changes the motion vector restriction parameter locally at slice boundaries based on scene characteristics. The system analyzes temporal complexity in different regions and adjusts the maximum motion vector value accordingly. This localized parameter adaptation resolves the contradiction by maintaining image quality at boundaries where it matters most while preserving overall processing efficiency through screen division encoding.
Data Source
AI summary
In a video encoding device, a transmission section sets a value corresponding to characteristics specified by ITU-R BT.709 to 1 in the transfer characteristics syntax when the dynamic range of the video signal is a SDR (Standard Dynamic Range), and sets a value corresponding to one of characteristics specified by HLG (Hybrid Log Gamma) to 18 or a value corresponding to one of characteristics specified by PQ (Perceptual Quantizer) to 16 when the dynamic range of the video signal is a HDR (High Dynamic Range).