ROIC Lag Correction via Filter Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Imaging devices face challenges in correcting image lag or persistence due to incomplete reset of pixels and bandwidth constraints, which historically made digital correction prohibitive due to physical constraints of the silicon area in readout circuits.
Innovation Solution
A method involving a readout interface circuit (ROIC) with a processor and buffer that determines filter coefficients using lookup tables to correct pixel outputs by multiplying current and prior frame pixel values, effectively reducing image lag by generating a more accurate representation of incident illumination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If digital correction is implemented to remove image lag, then image data accuracy is improved, but silicon area requirements increase making correction prohibitive
Solution Approach 1:
The patent merges the lag correction functionality directly into the ROIC by integrating the buffer and processor within the same chip structure. This combines multiple functions (pixel readout, data buffering, filter coefficient retrieval, and lag correction calculation) into a single integrated circuit, achieving effective lag correction without proportionally increasing overall silicon area usage.
Solution Approach 2:
The patent transitions from traditional external correction methods to an integrated on-chip solution by adding temporal processing dimensions. It buffers multiple frames temporally and applies filter coefficients across time dimensions, enabling sophisticated lag correction while maintaining spatial efficiency through the integrated architecture.
2Speed
If buffer and processor are integrated in ROIC for real-time lag correction, then correction speed is improved, but device complexity increases
Solution Approach 1:
The patent segments the lag correction process into distinct functional modules within the ROIC: a buffer for storing pixel values from multiple frames, a lookup table for filter coefficients, and a processor for executing the correction calculation. This modular segmentation enables real-time processing while managing complexity through organized, separable functional blocks.
Solution Approach 2:
The patent introduces a lookup table as an intermediary structure that stores pre-computed filter coefficients. This intermediary allows the processor to quickly retrieve appropriate coefficients without complex real-time calculations, speeding up the correction process while keeping the processor's computational burden manageable.
3Measurement precision
If multiple frame buffering is implemented, then lag correction accuracy is improved, but memory requirements increase
Solution Approach 1:
The patent implements local quality by creating pixel-specific correction paths within the ROIC. Each pixel's data is buffered and processed individually with its own filter coefficients, allowing precise lag correction tailored to each pixel's characteristics while efficiently using memory resources through localized processing rather than global buffering of all pixel data.
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
This approach systematically removes image lag in real-time without requiring external resources, allowing for improved image data accuracy and extending the use of lower bandwidth imaging pixel architectures, while reducing read noise.
Implementation Method 1
The sensor typically converts light incident on the sensor into a photocurrent
Data Source
AI summary
A method of correcting lag in an imaging pixel includes receiving a current frame pixel value and determining a current filter coefficient using the current frame pixel value. A pixel output is determined from a product of the current frame pixel value and current frame filter coefficient. The product of a first prior frame pixel value and corresponding first prior frame filter coefficient is added to the pixel output to generate a corrected pixel output to more closely indicates incident illumination on the imaging pixel during an integration period from which the current frame pixel value was obtained.