RGBW Imaging Sensor False Color Correction via Luminance Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In imaging devices with RGBW color filter arrangements, false color artifacts occur due to luminance variations and noise, particularly at lower light levels, affecting image quality and accuracy.
Innovation Solution
An imaging device with a signal processing unit that includes a luminance signal processing unit to generate luminance values from higher sensitivity white pixels and a false color determination unit to identify false color by comparing these values to a threshold, enabling accurate detection and correction of color errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a CF of RGBW arrangement is used to improve detection sensitivity, then sensitivity is improved, but false color is likely to occur due to lower density of RGB pixels
Solution Approach 1:
The patent performs preliminary classification of pixels into high-luminance and low-luminance groups before demosaicing, and selectively applies different processing methods to each group. This preliminary action allows the system to handle low-luminance pixels (prone to false color) differently from high-luminance pixels, thereby preventing false color while maintaining the sensitivity benefits of RGBW arrangement.
Solution Approach 2:
The patent dynamically changes the demosaicing method based on the luminance parameter of each pixel. For low-luminance pixels, it uses a method that references multiple surrounding pixels to calculate color values, while for high-luminance pixels, it uses standard demosaicing. This parameter-based adaptation resolves the contradiction by adjusting processing intensity according to actual pixel conditions.
2Loss of information
If demosaicing is performed based on pixel value ratio of RGB and W pixels, then color information is generated, but error from offset component and random noise causes false color at lower luminance
Solution Approach 1:
The patent introduces luminance information from W pixels as an intermediary to guide the demosaicing process. Instead of directly using noisy RGB pixel values at low luminance, the system uses the high-sensitivity W pixel luminance data to determine processing strategy and to assist in calculating accurate color values, thereby reducing the impact of offset and noise errors.
Solution Approach 2:
The patent implements a feedback mechanism where the luminance value of each pixel (from W pixels) is used to control the demosaicing process for corresponding RGB pixels. The system continuously adapts its processing based on the measured luminance level, creating a closed-loop control that prevents false color by adjusting demosaicing intensity according to actual signal quality.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution effectively determines and corrects false color artifacts in images from RGBW arrangements, improving image quality and accuracy by accurately identifying and addressing color errors caused by luminance variations and noise.
Implementation Method 1
an imaging element 100 that includes a plurality of pixels 12
Data Source
AI summary
A disclosed imaging device includes an imaging element that includes a plurality of pixels including a plurality of first pixels each of which outputs a signal including color information and a plurality of second pixels each of which has higher sensitivity than the first pixels, and a signal processing unit that processes a signal output from the imaging element. The signal processing unit includes a luminance signal processing unit that generates luminance values of the first pixels based on signals output from the second pixels and a false color determination unit that determines a presence or absence of false color based on a result of comparison between the luminance values of the first pixels generated by the luminance signal processing unit and a predetermined threshold value.