Image Sensor Column Gain Correction Using Dedicated Test Pixels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
CMOS image sensors face challenges in correcting column gain mismatches, particularly in low-light conditions where coherent noise artifacts become visible, and existing digital correction methods are insufficient to mask these mismatches under all lighting scenarios.
Innovation Solution
Incorporating at least one test signal pixel per column, where the test signal is compared to a target value to calculate and store correction values, which are then applied to regular photo-pixel signals to adjust for column gain mismatches, using a digital memory element and logic block to facilitate continuous adjustments and optimal memory utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If digital correction methods are applied to correct column gain mismatches, then image quality improves by reducing visible noise artifacts, but the correction is insufficient to mask mismatches under all lighting scenarios particularly in low-light conditions
Solution Approach 1:
The patent applies preliminary action by measuring and storing correction values for each column during a calibration phase before actual image capture. The system pre-determines the gain mismatch characteristics and stores correction coefficients in memory, so that when images are captured, the corrections are already prepared and can be applied immediately to mask column FPN artifacts effectively across all lighting conditions.
Solution Approach 2:
The patent employs parameter changes by adjusting the correction values stored in memory based on different operating conditions. The system measures column FPN characteristics under specific lighting conditions and stores correction parameters that can be applied to compensate for gain mismatches. By changing the correction parameters based on measured characteristics, the system adapts to different lighting scenarios and improves correction reliability.
2Measurement precision
If test signal pixels are added to each column for measuring gain mismatch, then correction accuracy improves, but device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the sensor array into functional segments: regular photo-capturing pixels and dedicated test signal pixels. Each column contains at least one test signal pixel that is specifically designed for measuring gain mismatch characteristics. This segmentation allows the system to perform accurate measurements without requiring all pixels to be complex test pixels, thus improving measurement precision while controlling overall device complexity.
Solution Approach 2:
The patent employs universality by designing test signal pixels that serve multiple functions: they capture test signals for gain mismatch measurement, and their correction values are applied universally to correct all pixels in the corresponding column. This multi-functionality allows a small number of dedicated test pixels to provide correction benefits for the entire column, improving measurement accuracy without proportionally increasing device complexity.
Data Source
Figure 1
Figure 2
Figure 3(A)~3(B)
AI summary
An apparatus and method for digital column gain mismatch correction are described. At least one test signal pixel is added to each column in an image sensor. For each column the test signal from at least one test pixel is read and compared to a target value. A correction value is calculated based on comparing the test signal to the target value. The correction value is stored in a memory. When the regular photo-pixels are read, the photo-pixel signals are modified based on the stored correction value for each column.