Image Encoding Segmentation for Error Resilience

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image encoding methods lack robustness and flexibility in handling errors during wireless transmission, particularly in maintaining high-quality image transmission while reducing data size, and do not efficiently adapt to different image modalities.

Innovation Solution

The method involves splitting images into independent portions, applying a frequency-based transform, quantizing coefficients, and converting them into binary code, with options for varying block sizes and quantization levels, and using pre-filters and probability models tailored to specific image modalities to enhance resilience and compression efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the image is processed as a single unit using traditional encoding methods, then the encoding process is simpler, but errors during transmission can propagate through the entire image and the processing speed is limited

Engineering Contradiction:
Improveerror resilienceVSAvoidencoding structure
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The image is divided into multiple independent image portions, where each portion is processed separately through the encoding pipeline. This segmentation ensures that errors during transmission are confined to individual portions rather than propagating across the entire image, thereby improving error resilience while maintaining a manageable encoding structure

Inventive Principle:
Principle #1Segmentation

2Productivity

If the entire image is encoded using a single thread, then the encoding process is simpler to implement, but the processing speed is slower

Engineering Contradiction:
Improveencoding speedVSAvoidprocessing architecture
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The image is divided into multiple independent portions that can be processed simultaneously using multi-threaded implementations. Each thread handles a separate image portion through the complete encoding pipeline, enabling parallel processing that significantly improves encoding speed while keeping each thread's logic relatively simple

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The encoding system dynamically adapts to different processing requirements by allowing flexible configuration of the number of threads and image portions. This dynamic architecture enables the system to optimize processing speed based on available computational resources while maintaining implementation simplicity

Inventive Principle:
Principle #15Dynamics

3Adaptability or versatility

If a uniform block size is used for all image portions, then the encoding process is more consistent and simpler, but the flexibility to adapt to different image regions and modalities is reduced

Engineering Contradiction:
Improveblock size adaptationVSAvoidencoding configuration
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

Different block sizes can be assigned to different image portions based on their specific characteristics and requirements. This local adaptation allows the encoding process to optimize for regions with sharp edges using smaller blocks while using larger blocks for uniform regions, improving overall encoding efficiency without requiring complex per-block configuration

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The encoding system provides a universal framework that can handle both uniform and non-uniform block size requirements through a single configurable interface. The same encoding pipeline can process image portions with different block sizes, making the system versatile across different image modalities and applications without requiring separate encoding configurations

Inventive Principle:
Principle #6Universality (Multi-functionality)

4Manufacturing precision

If a single quantisation level is used for the entire image, then the encoding process is simpler and more consistent, but the ability to capture high resolution in important regions while maintaining compression in other regions is lost

Engineering Contradiction:
Improvequantisation resolutionVSAvoidquantisation control
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

Different quantisation levels can be applied to different image portions based on their importance and characteristics. Regions of interest can be encoded with finer quantisation levels to preserve high resolution details, while less important regions use coarser quantisation levels to achieve higher compression ratios, all within a single configurable encoding process

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP4294017A1Method for image encoding
Publication Date: 2023.12.20 MBDA UK
  • EP4294017A1 patent drawingFigure 1a~1B
  • EP4294017A1 patent drawingFigure 2~6
  • EP4294017A1 patent drawingFigure 7~8

AI summary

A method for encoding data defining an image is disclosed. The image is split into a number of image portions. Each portion is segmented into image blocks, each image block in the portion having a uniform block size. A frequency-based transform is applied to each of the image blocks, thereby providing transformed image data in which the image data is represented as coefficients defining a linear combination of predetermined basis functions having different spatial frequencies. The coefficients are quantised, and converted into binary code. Each of the image portions being processed independently of the other image portions.