Image Sensor Black Level Stabilization via Analog-Digital Noise Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced imaging sensors face challenges in suppressing fixed pattern noise, particularly line and column noise, which degrade image quality and introduce frame-to-frame instability, especially at low illumination levels, due to inherent noise in common reference voltages and inefficient noise correction methods.
Innovation Solution
A method and circuit implementation that jointly utilize analog and digital circuits, along with optimized algorithms, to suppress fixed pattern noise by calculating and applying noise offsets on a line-by-line and column-by-column basis, using black clamp blocks, line noise correction blocks, and column noise correction blocks, while updating noise coefficients differentially to stabilize the black level and maximize dynamic range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If common reference voltages are used in column buffers, then circuit simplicity is improved, but noise and frame-to-frame instability are introduced
Solution Approach 1:
The patent divides the reference voltage into separate dedicated black reference pixels distributed across multiple rows, isolating the noise source from the signal path. Each black reference pixel operates independently to measure and correct black level variations, eliminating the frame-to-frame instability caused by shared reference voltages.
Solution Approach 2:
The patent introduces black reference pixels as intermediary elements that measure the actual black level at each location and time. These reference pixels serve as mediators between the ideal black level and the actual sensor output, enabling dynamic correction of noise and offset variations without affecting the main signal path.
2Manufacturing precision
If analog domain offset correction is applied, then image quality is improved, but processing speed and dynamic range are compromised
Solution Approach 1:
The patent performs offset correction in the digital domain after analog-to-digital conversion, using pre-measured black reference values to calculate and subtract offsets. This preliminary digital processing eliminates the need for complex analog correction circuits, maintaining processing speed while improving image quality through precise digital subtraction.
3Reliability
If black reference pixels are read during vertical blanking time, then black level stabilization is improved, but frame rate is reduced
Solution Approach 1:
The patent reads black reference pixels during vertical blanking intervals when no image data is being transmitted, utilizing otherwise wasted time. This partial utilization of blanking time for reference measurement enables black level stabilization without significantly impacting the main image frame rate, as the correction is applied digitally during processing.
Data Source
AI summary
A black clamp stabilization circuit for an image sensor utilizes a mixed-signal SoC block comprising sub-blocks to dynamically and precisely adjust the black level based on comparison to a reference black level. The black level adjustments include a first level regulation using digital control of an analog signal in a feedback loop that includes a programmable gain amplifier and high-resolution A/D converter. By applying the black clamping in the analog domain, dynamic range is extended. Additional black level regulation is subsequently performed in the digital domain to differentially eliminate line noise and column noise generated within the imaging System-on-Chip. By providing information between the sub-blocks, the algorithms can converge more quickly. The technique enables multiple signal paths to separately handle individual colors and to increase imaging data throughput.