Video Signal Encoding with Segmented Background Resolution

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing N-tiles format for encoding multiple 3D video signals into a single signal results in significant degradation of resolution, leading to blurred images and prominent artifacts, especially in stationary or slowly varying scenes, as it downsamples each channel to fit into existing 2D video chains.

Innovation Solution

The method separates video signals into non-overlapping subsequences based on scene characteristics, integrating them using edge-preserved low pass filters for stationary content and maintaining full resolution, while using the N-tiles format for rapidly changing areas, allowing for interlacing of background and foreground images to preserve high definition in static scenes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If downsampling is applied to fit multiple video channels into existing 2D video chains, then compatibility with existing video chains is improved, but resolution is significantly degraded

Engineering Contradiction:

Improvecompatibility with existing video chainsVSAvoidresolution

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The video signal is segmented into background and foreground components based on motion detection. Background regions (stationary or slow-moving) are extracted and processed separately from foreground regions (fast-moving), allowing different resolution treatments for different parts of the image.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different quality levels are applied to different regions of the video signal. Background regions are maintained at full resolution while foreground regions are downsampled, creating a spatially varying quality distribution that optimizes both visual perception and bandwidth utilization.

Inventive Principle:

Principle #3Local quality

2Productivity

If downsampling is applied to encode multiple channels into a single video signal, then the ability to distribute and compress video signals is improved, but picture quality is heavily degraded

Engineering Contradiction:

Improveability to distribute and compress video signalsVSAvoidpicture quality

Core Design Contradiction:

ProductivityVSManufacturing precision

Solution Approach 1:

The video content is segmented into background and foreground based on motion characteristics. This segmentation enables selective processing where background regions maintain full resolution for high quality, while foreground regions are downsampled to enable efficient compression and distribution.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The resolution parameter is dynamically changed based on motion activity. Background regions undergo minimal downsampling or no downsampling, while foreground regions are heavily downsampled. This parameter adaptation allows the system to achieve good compression ratios while preserving visual quality in static regions.

Inventive Principle:

Principle #35Parameter changes

3Manufacturing precision

If interpolation techniques are applied to recover detail information, then visual quality is improved, but computational loading is heavily increased

Engineering Contradiction:

Improvevisual qualityVSAvoidcomputational loading

Core Design Contradiction:

Manufacturing precisionVSPower

Solution Approach 1:

Motion detection and background extraction are performed as preliminary actions before compression and transmission. By identifying and separating background regions in advance, the system avoids the need for heavy interpolation computations during playback, as background regions are already preserved at full resolution.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

Background regions are extracted from the video signal and handled separately from foreground regions. This extraction eliminates the need for interpolation on background areas, significantly reducing computational loading while maintaining visual quality in these regions.

Inventive Principle:

Principle #2Taking out (Extraction)

4Adaptability or versatility

If N-tiles format is used to integrate multi-view 3D video signals, then compatibility with existing video chains is improved, but resolution of each view is degraded

Engineering Contradiction:

Improvecompatibility with existing video chainsVSAvoidresolution of each view

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The N-tiles format is modified by segmenting each tile into background and foreground regions. Background regions are preserved at full resolution while foreground regions are downsampled, allowing the system to maintain compatibility with N-tiles infrastructure while improving overall visual quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different quality levels are applied within each N-tile based on local motion characteristics. Background regions across all tiles maintain full resolution, while only foreground regions undergo downsampling, creating a differentiated quality distribution that improves perceived image quality while maintaining compatibility.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS8175398B2Method for encoding a plurality of video signals into a single video signal

Publication Date: 2012.05.08 CITY UNIVERSITY OF HONG KONG

AI summary

A method of encoding video signals into a single video signal includes extracting background images from corresponding video signals, encoding the video signals into a single video signal, the single video signal containing information for reconstructing the video signals, and replacing frames of the single video signal with the background images.