Image Sensor Pixel Saturation Correction via Multi-Exposure Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image sensors using arrays of red, green, and blue pixels face inaccuracies in color reproduction under varying illumination conditions, particularly when substituting green pixels with high-sensitivity pixels like yellow, as they tend to saturate quickly in bright environments, leading to noise and inaccurate color representation.
Innovation Solution
An electronic apparatus and method that utilize an image sensor to acquire pixel values during different exposure times, with a controller correcting saturated pixel values by using pixel values from nearby pixels of the same color, thereby improving color reproducibility and reducing noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If yellow pixels are used to replace green pixels to reduce noise in low illumination environments, then noise reduction is improved, but pixel values rapidly reach maximum limit value in sufficient illumination environments causing color inaccuracy
Solution Approach 1:
The patent divides pixels into multiple groups with different exposure times. First pixels use a first exposure time while second pixels use a second exposure time (longer than the first). This segmentation allows different pixel groups to serve different illumination conditions, preventing saturation in bright environments while maintaining sensitivity in dark environments.
Solution Approach 2:
The patent changes the exposure time parameter for different pixel groups. By adjusting exposure times as a variable parameter, the system optimizes both noise reduction and color accuracy. The controller corrects pixel values based on exposure time ratios, dynamically adapting to illumination conditions.
2Illumination intensity
If longer exposure time is used to improve light capture in low illumination environments, then sensitivity is improved, but pixel values saturate quickly in sufficient illumination environments
Solution Approach 1:
The patent segments pixels into first pixels with shorter exposure time and second pixels with longer exposure time. This allows the system to capture light effectively in low illumination conditions while preventing saturation in bright conditions through the shorter exposure group.
Solution Approach 2:
The patent applies partial action by using longer exposure time only for specific pixel groups (second pixels) rather than all pixels. This partial application of long exposure achieves enhanced light capture where needed without causing universal saturation.
3Measurement precision
If multiple exposure times are used to correct saturated pixels, then color reproducibility is improved, but device complexity increases
Solution Approach 1:
The patent makes the exposure time system multi-functional by using the same hardware for both short and long exposure modes. The controller manages different exposure times through software control, making the system versatile for various illumination conditions without requiring separate hardware for each mode.
Solution Approach 2:
The controller corrects pixel values based on feedback from saturation detection. By comparing pixel values against threshold values and using the ratio of exposure times, the system dynamically adjusts and corrects saturated pixel values to maintain color accuracy.
Data Source
AI summary
An electronic apparatus includes: an image sensor for acquiring pixel values of first pixels sensed during a first exposure time and second pixels sensed during a second exposure time longer than the first exposure time; and a controller for outputting an output image acquired based on pixel values of the first pixels and a corrected saturated pixel value obtained by correcting a pixel value of a saturated pixel having a pixel value exceeding a threshold value among the second pixels, using a pixel value of at least one first pixel having a distance closest to a position of the saturated pixel among the first pixels.