Adaptive Local White Balance Adjustment for CMOS Image Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic white balance methods in CMOS image sensors fail to provide proper color correction in scenes illuminated by multiple light sources, often resulting in incorrect color casts, especially in outdoor scenes with shadows.
Innovation Solution
An adaptive local white balance adjustment method that combines anisotropic filtering and bilateral filtering to calculate weights for each pixel, using both local and global components to preserve chromaticity and prevent saturation, allowing for edge-aware and color-accurate adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If common AWB methods are applied to scenes with multiple illuminants, then a trade-off is made between different illuminants, but this results in incorrect color casts in both parts of the image
Solution Approach 1:
The image is segmented into multiple regions based on illuminant characteristics. The patent identifies different illuminant regions (e.g., sunlight areas, shadow areas) and applies separate white balance calculations to each region, allowing accurate color correction for each illuminant type rather than forcing a single global white balance setting.
Solution Approach 2:
The patent applies local white balance adjustment by calculating white balance parameters independently for different regions of the image. Each region's white balance is determined based on its local illuminant characteristics, enabling precise color accuracy in each area while maintaining overall image coherence through the use of weighting functions.
2Device complexity
If global white balance adjustment is applied to the entire image, then processing is simplified, but this causes loss of local chromaticity information and saturation
Solution Approach 1:
The patent implements a dynamic white balance system that adapts its processing approach based on local image characteristics. By using spatial weighting functions and region-based analysis, the system dynamically adjusts white balance parameters for different areas, preserving local chromaticity information while maintaining computational feasibility through structured processing algorithms.
Solution Approach 2:
The patent introduces intermediate processing steps including region identification, local white balance calculation, and weighted combination of local and global adjustments. These intermediary processes act as mediators between simple global adjustment and complex pixel-by-pixel processing, preserving chromaticity information while controlling computational complexity.
Data Source
AI summary
The disclosure describes embodiments of an apparatus comprising an image sensor including a pixel array having a plurality of pixels and an automatic white balance (AWB) circuit coupled to the pixel array. The AWB circuit is used to determine a local white balance component for each pixel, determine a global white balance component for each pixel, and apply a white balance adjustment to each pixel, the applied white balance adjustment comprising a combination of the local white balance component and the global white balance component. The disclosure also describes embodiment of a process including receiving image data from each pixel in a pixel array, determining a local white balance component for the image data from each pixel, determining a global white balance component for the image data from each pixel, and applying a white balance adjustment to the image data from each pixel, the applied white balance adjustment comprising a combination of the local white balance component and the global white balance component. Other embodiments are also disclosed and claimed.