Vehicle Camera Image Processing for Bandwidth-Constrained Transmission

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The limited bandwidth for wireless transmission of high-definition video data from vehicle-mounted cameras poses a challenge, as existing technologies struggle to efficiently transmit video data while maintaining image quality, especially with wide-angle lenses, due to the high data volume and cost associated with advanced compression techniques.

Innovation Solution

The method involves performing scene analysis to identify regions of greater and lesser interest, processing the image data to reduce information in lesser interest areas, and compressing the data differently for efficient transmission, using techniques like filtering and edge detection, with machine learning algorithms for accurate analysis and enhancement of regions of interest.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If advanced data compression techniques are used to transmit high-definition video data, then transmission cost is reduced, but transmission bandwidth requirements still exceed available wireless bandwidth

Engineering Contradiction:
Improvetransmission costVSAvoiddata volume
Core Design Contradiction:
Loss of energyVSQuantity of substance

Solution Approach 1:

The image is divided into multiple regions of interest (ROIs) and corresponding non-ROI regions. Different compression techniques and compression ratios are applied to different regions, allowing efficient transmission by focusing high compression on non-critical areas while maintaining quality in important regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different quality levels are applied to different regions of the image. Non-ROI regions undergo aggressive compression with lower quality requirements, while ROI regions maintain high definition quality. This local differentiation reduces overall data volume while preserving essential information.

Inventive Principle:
Principle #3Local quality

2Area of stationary object

If wide angled lenses are used to capture broad pictures, then field of view is increased, but data volume increases making transmission difficult

Engineering Contradiction:
Improvefield of viewVSAvoiddata volume
Core Design Contradiction:
Area of stationary objectVSQuantity of substance

Solution Approach 1:

The wide-angle image is segmented into multiple regions of interest based on scene analysis. By identifying and isolating specific ROIs within the broad field of view, the system can apply selective compression to non-ROI areas, reducing overall data volume while maintaining comprehensive scene coverage.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system extracts and prioritizes transmission of critical region information while reducing or omitting less important background areas. This extraction approach allows the wide-angle lens to capture comprehensive scenes without transmitting all captured data at full resolution.

Inventive Principle:
Principle #2Taking out (Extraction)

3Quantity of substance

If image data is compressed to reduce data size, then transmission bandwidth requirements are reduced, but image quality deteriorates

Engineering Contradiction:
Improvedata sizeVSAvoidimage quality
Core Design Contradiction:
Quantity of substanceVSLoss of information

Solution Approach 1:

The system applies different compression quality levels to different regions based on their importance. ROI regions maintain high definition quality with minimal compression, while non-ROI regions undergo aggressive compression. This local quality differentiation preserves essential image information while significantly reducing overall data size.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

Scene analysis is performed in advance to identify regions of interest before compression. This preliminary identification allows the compression algorithm to pre-plan which areas require high quality preservation and which can be aggressively compressed, optimizing the balance between data size and image quality.

Inventive Principle:
Principle #10Preliminary action

4Productivity

If information is reduced in regions of lesser interest, then compression efficiency is improved, but information completeness is reduced

Engineering Contradiction:
Improvecompression efficiencyVSAvoidinformation completeness
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The system selectively reduces information quality in non-ROI regions while maintaining full information integrity in ROI regions. This local differentiation improves compression efficiency by allowing aggressive compression where information can be reduced without impacting critical details, while preserving information completeness in important areas.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system applies partial compression to non-ROI regions, reducing information only to the extent necessary for efficient transmission. Critical information in these regions is preserved while redundant or less important details are reduced, achieving a balance between compression efficiency and information completeness.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3975133A1Processing of images captured by vehicle mounted cameras
Publication Date: 2022.03.30 TRAKM8
  • EP3975133A1 patent drawingFigure 1
  • EP3975133A1 patent drawingFigure 2
  • EP3975133A1 patent drawingFigure 3

AI summary

A method of processing images captured by an on-vehicle camera, circuitry and the on-board camera are disclosed The method comprises: performing scene analysis on image data received from the on-vehicle camera to identify regions of greater interest and regions of lesser interest in images in the image data; processing the image data to preferentially reduce the information in the regions of lesser interest to generate processed image data; and compressing at least a portion of the processed image data to generate compressed image data.