Column Parallel Image Sensor Calibration via Offset Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
CMOS semiconductor imagers with column parallel architecture face challenges in calibrating plural pixel signal chains due to variations inherent in semiconductor fabrication, leading to noise in images produced.
Innovation Solution
Incorporating a calibration row of calibration pixels within the pixel array, which provides offset values that can be applied to digital values from imaging pixels to compensate for differences in signal chain responses, using column circuitry to sample and process signals from both conventional and calibration pixels to derive and apply calibration values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a conventional pixel array without calibration pixels is used, then the device complexity is reduced, but the manufacturing precision and image quality deteriorate due to uncorrected signal chain variations
Solution Approach 1:
The pixel array is segmented into functional regions: imaging pixels for capturing images and calibration pixels for deriving offset values. This segmentation allows independent optimization of each region's function while maintaining overall system integrity.
Solution Approach 2:
Calibration pixels are created as copies of imaging pixels with identical signal chain architecture. These copies reproduce the same signal processing characteristics but without the need for actual image capture, enabling offset value derivation through comparison with known reference values.
2Measurement precision
If calibration pixels are added to the pixel array, then the manufacturing precision improves through signal chain calibration, but the device complexity increases
Solution Approach 1:
The calibration pixels enable the system to self-calibrate by deriving offset values from the calibration pixel outputs and applying them to imaging pixel signals. This self-service mechanism eliminates the need for external calibration equipment or manual adjustment processes.
Solution Approach 2:
The system uses calibration pixel outputs as feedback to derive offset values, which are then fed back into the signal processing path to correct imaging pixel signals. This feedback loop continuously improves measurement precision by compensating for signal chain variations.
3Productivity
If multiple signal chains are used in column parallel architecture, then the productivity increases through parallel processing, but the reliability decreases due to variations between signal chains
Solution Approach 1:
Offset values are derived as correction parameters that adjust the signal chain responses. By changing the electrical parameters of the signal processing path through these offset corrections, the system maintains consistent signal chain behavior across all parallel chains while preserving their parallel processing capability.
Solution Approach 2:
The calibration process equalizes the signal chain responses by applying appropriate offset values to each chain. This creates equipotentiality in terms of signal chain performance, ensuring all chains operate at the same level of accuracy and consistency while maintaining parallel processing efficiency.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The plural signal chains of an imaging device are calibrated in the digital domain. The pixel array of the imaging device includes a row of calibration pixels. The column circuitry, prior to reading a row of pixels, reads the row of calibration pixels in order to obtain a set of column offsets. The row of pixels is read and processed to produce a corresponding plurality of digital values. The set of column offsets are applied to the digital values to compensate for response differences among the signal chains.