Histogram Segmentation Adaptive Wiener Filter for Video Encoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In lossy video coding, quantization of transform coefficients leads to quality degradation in reconstructed pictures, and global Wiener filters may fail to adapt to local pixel information and constant quality loss within histogram bins.

Innovation Solution

Adaptive Wiener filtering with offset is applied to histogram bins, using histogram segmentation to categorize pixels and determine the best filtering scheme based on Rate Distortion Optimization (RDO) criteria, with filter coefficients and offset values generated and transmitted to the decoder to minimize sum of absolute differences between original and filtered pixel values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If a global Wiener filter is applied to reconstructed pictures, then picture quality improvement is achieved, but the filter fails to adapt to local pixel information and constant quality loss within histogram bins

Engineering Contradiction:
Improvepicture qualityVSAvoidadaptation to local pixel information
Core Design Contradiction:
Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The patent segments the histogram into multiple bins based on pixel value ranges, allowing different filtering strategies to be applied to different segments. This segmentation enables the filter to adapt to local characteristics within each bin while maintaining computational efficiency through batch processing of segmented data.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies local adaptive filtering within each histogram bin by calculating offset values specific to each bin. This allows the filter to maintain different quality characteristics for different pixel value ranges, addressing the constant quality loss problem within bins while improving overall picture quality through localized adaptation.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If adaptive filtering is applied to compensate for picture quality loss, then reconstructed picture quality improves, but computational complexity and processing overhead increase

Engineering Contradiction:
Improvereconstructed picture qualityVSAvoidfiltering processing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary histogram analysis and bin segmentation before applying the Wiener filter. By pre-calculating histogram bins and organizing pixel data into segments, the complex adaptive filtering operation is simplified into more manageable batch operations, reducing overall computational complexity while maintaining adaptation benefits.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent changes the parameter representation by working with histogram bin offsets rather than individual pixel values. This parameter transformation allows the filter to adapt to local characteristics using compact offset values for each bin, significantly reducing the computational burden compared to processing each pixel individually while maintaining picture quality improvement.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If quantization step is increased to improve video compression gain, then compression efficiency improves, but picture quality loss increases

Engineering Contradiction:
Improvevideo compression gainVSAvoidpicture quality
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The patent converts the harmful effect of quantization-induced quality loss into a benefit by using the Wiener filter to statistically reconstruct lost information. The filter exploits the statistical properties of quantized data to recover picture quality, effectively turning the compression artifact into an opportunity for intelligent reconstruction and quality improvement.

Inventive Principle:
Principle #22Blessing in disguise (Convert harm into benefit)

Solution Approach 2:

The patent implements a feedback mechanism where the reconstructed picture from the Wiener filter is used to improve subsequent encoding decisions. By analyzing the filtering results and their impact on quality, the system can adjust encoding parameters to maintain better quality at comparable compression ratios, creating a continuous improvement loop.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3242476B1Histogram segmentation based local adaptive filter for video encoding and decoding
Publication Date: 2020.10.21 INTEL CORP
  • EP3242476B1 patent drawingFigure 1~2
  • EP3242476B1 patent drawingFigure 3
  • EP3242476B1 patent drawingFigure 4

AI summary

Reconstructed picture quality for a video codec system may be improved by categorizing reconstructed pixels into different histogram bins with histogram segmentation and then applying different filters on different bins. Histogram segmentation may be performed by averagely dividing the histogram into M bins or adaptively dividing the histogram into N bins based on the histogram characteristics. Here M and N may be a predefined, fixed, non-negative integer value or an adaptively generated value at encoder side and may be sent to decoder through the coded bitstream.