Sub-Image Matching in Video Encoding for Desktop Bandwidth Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Video encoding in desktop scenarios, such as page scrolling and page flipping, is bandwidth-intensive and inefficient due to repeated encoding of identical image areas.

Innovation Solution

An image encoding method that partitions images into sub-images, classifies them based on feature values, and reuses encoding data for matched sub-images, while performing standard encoding on unmatched sub-images, reducing repeated encoding of valid pixels and optimizing bandwidth usage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If video encoding reuses image encoding data from adjacent previous frames for current frame, then encoding efficiency is improved, but bandwidth consumption increases in desktop scenarios

Engineering Contradiction:
Improveencoding efficiencyVSAvoidbandwidth consumption
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The patent divides the image into multiple sub-images and processes each sub-image independently. By segmenting the encoding task at the sub-image level rather than frame level, the system can identify and reuse only the specific sub-images that remain unchanged, rather than reusing entire frame data. This granular segmentation enables precise bandwidth optimization while maintaining encoding efficiency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different encoding strategies to different sub-images based on their change status. Sub-images that remain unchanged reuse encoding data from previous frames, while changed sub-images undergo full encoding. This local differentiation optimizes bandwidth consumption by transmitting only necessary data while maintaining overall encoding efficiency.

Inventive Principle:
Principle #3Local quality

2Reliability

If video encoding processes all sub-images using standard encoding, then encoding completeness is ensured, but encoding efficiency decreases due to repeated encoding of identical content

Engineering Contradiction:
Improveencoding completenessVSAvoidencoding efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent performs preliminary comparison between current sub-images and previous frame sub-images before encoding. By identifying unchanged sub-images in advance, the system can skip the encoding process for these regions and directly reuse previous encoding data. This preliminary action ensures that only necessary encoding operations are performed, maintaining completeness while improving efficiency.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent copies encoding data from previous frames for sub-images that remain unchanged. Instead of re-encoding identical content, the system directly copies the existing encoding data, ensuring encoding completeness for unchanged regions while significantly improving overall encoding efficiency by avoiding redundant processing.

Inventive Principle:
Principle #26Copying

3Measurement precision

If video encoding compares current frame with adjacent previous frame, then motion detection is achieved, but bandwidth-intensive processing occurs in desktop scenarios with frequent page scrolling

Engineering Contradiction:
Improvemotion detection accuracyVSAvoidbandwidth intensity
Core Design Contradiction:
Measurement precisionVSLoss of energy

Solution Approach 1:

The patent segments the motion detection process to operate at the sub-image level rather than entire frame level. By dividing images into sub-images and comparing only corresponding sub-images between frames, the system maintains motion detection accuracy for changed regions while reducing overall bandwidth-intensive processing in desktop scenarios with frequent page scrolling.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4694119A1Image coding/decoding method and apparatus
Publication Date: 2026.02.11 HUAWEI TECH CO LTD
  • EP4694119A1 patent drawingFigure 1
  • EP4694119A1 patent drawingFigure 2
  • EP4694119A1 patent drawingFigure 3

AI summary

An image encoding method and apparatus are disclosed, and relate to the field of video processing technologies. An encoding device classifies sub-images of a to-be-encoded image based on feature values of the sub-images. Different sub-images have different feature values, and after matching the sub-images with encoded sub-images in a buffered queue, a matched sub-image and an unmatched sub-image may be determined. Therefore, image encoding data of a sub-image that is in a previous frame of image and that is the same as the matched sub-image in the to-be-encoded image can be reused only when the matched sub-image in the to-be-encoded image and the sub-image in the previous frame of image do not need to have a same location and same content. Further, in a process of encoding the to-be-encoded image, encoding (for example, standard encoding) is performed on the unmatched sub-image, and information about the matched encoded sub-image is generated for the matched sub-image. This avoids a problem of repeated encoding of valid pixels in the matched sub-images, reduces bandwidth for video encoding, and improves encoding efficiency.