Adaptive Film Grain Decoding for Consistent Picture Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing film grain modeling and synthesis methods in video encoding and decoding are not adaptable to different types of film grain, leading to inconsistent visual effects across varying resolutions and regions within a frame, and fail to consider film grain characteristics, resulting in suboptimal picture quality.
Innovation Solution
A decoding method that adapts film grain processing based on film grain adaptive indication information, determining parameters for film grain model, region, and layer to enhance visual effects by adding film grain to decoded pictures, and an encoding method that models film grain characteristics and writes adaptive indication information into the bitstream.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single film grain modeling and synthesis method is used, then the encoding and decoding process is simple, but the picture quality is poor and not suitable for different types of film grain
Solution Approach 1:
The patent implements dynamic film grain processing by enabling the decoder to adaptively select and apply different film grain modeling and synthesis methods based on the specific characteristics of the decoded picture. The system dynamically adjusts the film grain synthesis process according to picture resolution, region characteristics, and detected film grain properties, transforming a static single-method approach into a dynamic multi-method system that optimizes picture quality for varying conditions
Solution Approach 2:
The patent changes key parameters including picture resolution thresholds, region characteristics (texture complexity, brightness levels), and film grain properties (intensity, color temperature, sharpness). By monitoring these parameters and adjusting the film grain synthesis approach accordingly, the system achieves adaptability across different film grain types and picture conditions without requiring a completely complex restructured system
2Reliability
If film grain synthesis is applied to high-resolution pictures, then the effectiveness is high, but the effectiveness is low for low-resolution pictures
Solution Approach 1:
The patent applies local quality by treating high-resolution and low-resolution pictures differently. For high-resolution pictures, full film grain synthesis is applied to achieve realistic film grain effects. For low-resolution pictures, the system either skips film grain synthesis or applies a simplified version, recognizing that film grain effects are less perceptible and may even degrade quality in lower resolution contexts. This localized approach optimizes reliability for each resolution category
3Ease of operation
If a single film grain model is used for the entire frame, then the processing is uniform, but different visual effects are produced in different regions
Solution Approach 1:
The patent segments the picture frame into multiple regions based on characteristics such as texture complexity and brightness levels. Different film grain synthesis parameters and methods are applied to different regions - for example, regions with complex texture may use reduced film grain intensity to preserve detail, while uniform regions may use higher intensity for realistic film grain effects. This segmentation maintains operational simplicity through automated region classification while achieving precise visual consistency across diverse frame regions
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Provided are a decoding method, an encoding method, an electronic device, and a storage medium. The decoding method includes decoding a bitstream to obtain a decoded picture; acquiring film grain adaptive indication information corresponding to the decoded picture; determining film grain parameters corresponding to the decoded picture based on the film grain adaptive indication information; synthesizing film grain corresponding to the decoded picture based on the film grain parameters; and adding the film grain to the decoded picture.