Image Processing Device Automatic White Balance Block Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing Automatic White Balance (AWB) algorithms in digital cameras face challenges in accurately adjusting white balance, especially in diverse lighting conditions and when saturated pixels are present, leading to unreliable results.

Innovation Solution

The proposed solution involves dividing image frames into blocks, judging each block for potential greyness, excluding saturated pixels from calculations, and cumulating average values within defined intensity ranges to determine digital gains for white balance adjustment, using color temperature measures and sensor characteristics to enhance accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional AWB algorithms calculate averages for all color components, then the process is simple, but the accuracy deteriorates when saturated pixels are present

Engineering Contradiction:
Improvewhite balance accuracyVSAvoidalgorithm complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The image frame is divided into multiple blocks, and each block is individually evaluated for greyness and saturation. This segmentation allows the algorithm to process different regions with different characteristics, excluding saturated blocks from the white balance calculation while including grey blocks, thereby improving accuracy without requiring complex per-pixel analysis throughout the entire image.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different blocks are treated differently based on their local characteristics. Blocks judged to be grey and non-saturated are selected for white balance calculation, while saturated blocks are excluded. This local quality approach ensures that only reliable blocks contribute to the white balance gains, improving measurement precision.

Inventive Principle:
Principle #3Local quality

2Reliability

If all blocks are used for white balance calculation, then more data is available, but reliability deteriorates due to saturated pixels

Engineering Contradiction:
Improvewhite balance reliabilityVSAvoidnumber of blocks used
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

Blocks containing saturated pixels are extracted and excluded from the white balance calculation. The algorithm identifies and removes unreliable blocks (those with saturated pixels) from the set of blocks used for calculating white balance gains, ensuring that only reliable non-saturated blocks contribute to the final result.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The algorithm provides feedback by evaluating each block's saturation level and greyness characteristic, then uses this feedback to determine whether to include or exclude the block from white balance calculation. This feedback mechanism ensures that unreliable blocks are automatically filtered out, improving overall reliability.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If simple average calculation is used, then processing is fast, but accuracy deteriorates in diverse lighting conditions

Engineering Contradiction:
Improvewhite balance accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The image is segmented into blocks that are processed in parallel. By dividing the frame into multiple independent blocks and evaluating them separately, the algorithm can efficiently identify grey blocks and calculate white balance gains without requiring complex sequential processing, thus maintaining processing speed while improving accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Instead of processing all blocks equally, the algorithm performs partial action by selecting only the grey blocks for white balance calculation. This selective approach avoids unnecessary processing of saturated or non-representative blocks, reducing wasted computation time while improving the accuracy of the white balance adjustment.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP1920610B1Image processing device with automatic white balance
Publication Date: 2019.05.29 NOKIA TECHNOLOGIES OY
  • EP1920610B1 patent drawingFigure 1~2
  • EP1920610B1 patent drawingFigure 3
  • EP1920610B1 patent drawingFigure 4~5

AI summary

The purpose of the present invention is to provide sophisticated AWB technologies. According to one aspect of the present invention, there is provided a technology for adjusting a white balance of a frame of image data including a plurality of color elements. This technology is characterized by comprising: dividing the frame into a plurality of blocks including a plurality of pixel data; judging, for each of all of or a part of the blocks, whether the block is likely to be grey or not, and; deciding gains for adjusting a white balance using the blocks judged as being likely to be grey.