AVM Image Compositing Brightness Clustering

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional Around View Monitor (AVM) systems struggle to create a seamless composite image due to significant brightness differences between images from multiple cameras, leading to an unnatural viewing experience and difficulty in accurately correcting brightness errors.

Innovation Solution

A method using a clustering technique to minimize brightness differences by calculating representative brightness values, grouping similar curves, and applying correction values to each pixel, thereby smoothing the composite image and enhancing the driving experience.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If conventional image compositing methods are used to combine multiple camera images, then the composite image can be generated, but brightness differences between images cause unnatural appearance and reduce image quality

Engineering Contradiction:

Improvebrightness uniformityVSAvoidcompositing process complexity

Core Design Contradiction:

Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent segments the brightness correction problem by dividing images into multiple regions (sky region, ground region, and intermediate regions) and applying different correction strategies to each region. This allows precise control of brightness uniformity in each area while managing overall compositing complexity through systematic regional processing.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent changes brightness parameters through clustering analysis, grouping pixels with similar brightness characteristics together. By identifying representative brightness values for each cluster and adjusting these parameters systematically, the method achieves uniform brightness across composite images while maintaining a structured correction process.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If average brightness calculation is used to correct image brightness, then the correction process is simple, but errors occur due to shadows and other objects making accurate correction difficult

Engineering Contradiction:

Improvebrightness measurement accuracyVSAvoidbrightness correction complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies local quality by calculating brightness characteristics separately for different regions (sky, ground, intermediate areas) rather than using a global average. This regional approach ensures that shadows and objects in specific areas do not skew the brightness measurement for entire images, improving measurement precision while maintaining manageable correction complexity through localized processing.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent performs partial brightness correction by focusing on specific problematic regions (particularly intermediate regions between cameras) rather than attempting to correct entire images uniformly. This selective approach improves accuracy in critical areas without requiring excessive computational resources for complete image-wide correction.

Inventive Principle:

Principle #16Partial or excessive action

3Adaptability or versatility

If multiple cameras are installed to provide comprehensive vehicle surroundings view, then the monitoring coverage is improved, but brightness differences between cameras create unnatural composite images

Engineering Contradiction:

Improvesurroundings monitoring coverageVSAvoidcomposite image quality

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The patent achieves equipotentiality by equalizing brightness levels across all camera images through clustering-based correction. By adjusting brightness parameters so that corresponding regions in composite images have equal brightness characteristics, the method eliminates the unnatural appearance caused by multiple cameras while preserving the comprehensive monitoring coverage benefits.

Inventive Principle:

Principle #12Equipotentiality

4Manufacturing precision

If brightness correction is applied to all pixels in composite images, then the overall brightness uniformity is improved, but the processing time and computational load increase

Engineering Contradiction:

Improvebrightness uniformityVSAvoidimage processing time

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent segments the pixel population into distinct regions (sky, ground, intermediate regions) and processes only the most critical intermediate regions with full clustering-based correction. This selective segmentation approach maintains brightness uniformity in areas where it matters most while reducing overall processing time by limiting intensive computations to specific regions rather than applying uniform correction to all pixels.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS9367906B2System and method for compositing various images using clustering technique

Publication Date: 2016.06.14 HYUNDAI MOBIS CO LTD

AI summary

Disclosed are a system and a method for compositing various images that minimize a brightness difference in connection areas of various images by using a clustering technique at the time of compositing various images. The method for compositing various images may include receiving two or more input images; compositing the two or more input images into one composite image; calculating a brightness distribution degree; calculating representative brightness values; determining clustering; calculating a correction target value; and calculating a correction value of an increase/decrease curve.