Film Grain Estimation in Video Encoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional video encoding schemes struggle to efficiently compress video data containing film grain, often resulting in either high bitrate requirements to preserve grain quality or undesirable visual artifacts and temporal instability at reasonable bitrates.

Innovation Solution

A method that generates a decoded image by encoding the input image, determining noise model parameters from the decoded image, and including these parameters in the encoded stream, avoiding the use of denoised images to reduce computation complexity and leverage existing encoding operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If conventional video encoding schemes are used to preserve grain information, then grain quality is maintained, but bitrate requirement increases significantly

Engineering Contradiction:
Improvegrain qualityVSAvoidbitrate
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent extracts grain information from the video data by computing grain parameters based on the difference between the input video frame and the decoded video frame. This separation allows grain to be processed independently from the main video content, enabling efficient compression while preserving grain quality.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transforms grain representation from raw pixel data to parametric form by computing grain parameters (such as grain strength, color, and spatial characteristics). This parameterization reduces the data required to represent grain while maintaining its visual appearance, thereby reducing bitrate requirements.

Inventive Principle:
Principle #35Parameter changes

2Productivity

If conventional video encoding schemes are used with reasonable bitrate, then encoding efficiency is maintained, but grain information is lost or degraded with visual artifacts

Engineering Contradiction:
Improveencoding efficiencyVSAvoidgrain quality
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The patent performs preliminary grain parameter computation during the encoding process by analyzing the difference between input and decoded frames. This preliminary extraction of grain characteristics allows the encoder to allocate bits efficiently between grain parameters and video content, ensuring grain quality is preserved even at reasonable bitrates without compromising encoding efficiency.

Inventive Principle:
Principle #10Preliminary action

3Object-generated harmful factors

If denoised images are used for encoding, then noise is removed, but computation complexity increases

Engineering Contradiction:
Improvenoise removalVSAvoidcomputation complexity
Core Design Contradiction:
Object-generated harmful factorsVSDevice complexity

Solution Approach 1:

Instead of removing noise through complex denoising operations, the patent inverts the approach by directly modeling and parameterizing the noise (grain) characteristics from the encoded difference. This inversion simplifies computation by working with the residual grain signal rather than attempting to eliminate it through iterative denoising processes.

Inventive Principle:
Principle #13The other way round (Inversion)

Data Source

PatentEP4017011A1Method and apparatus for film grain estimation and coding
Publication Date: 2022.06.22 ATEME
  • EP4017011A1 patent drawingFigure 1
  • EP4017011A1 patent drawingFigure 2~3
  • EP4017011A1 patent drawingFigure 4

AI summary

A method of processing an image is proposed, which comprises: generating a decoded image by first encoding the image, wherein the first encoding the image comprises decoding encoded data generated based on the image, determining estimates of parameters of a parametric model of noise contained in the image based on the decoded image, including the estimates of parameters of the parametric model of noise in an encoded stream generated by second encoding the image.