Image Sensor Noise Suppression via Global Reference Array
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image sensor devices face challenges in effectively suppressing row-wise noise due to offsets or variations in non-imaging arrays, leading to artifacts in the final output image, and parasitic capacitance mismatches between imaging and non-imaging arrays that hinder perfect noise cancellation.
Innovation Solution
The solution involves an image sensor device with a global reference non-imaging array, separate control lines for imaging and reference arrays, and a non-imaging matching array to match parasitic capacitance, allowing for simultaneous sampling and averaging of reference values to subtract noise from image samples, thereby eliminating fixed pattern artifacts and improving noise rejection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate control lines are used for imaging and reference arrays, then noise cancellation accuracy is improved, but device complexity increases
Solution Approach 1:
The patent divides the control signal paths into separate segments: independent reset control lines (RST1, RST2) and independent transfer control lines (TX1, TX2) for the imaging array and reference array. This segmentation prevents noise coupling between the arrays while maintaining accurate noise cancellation by allowing each array to be controlled independently without interference from the other.
2Object-affected harmful factors
If non-imaging array is used for noise reference, then row-wise noise is suppressed, but parasitic capacitance mismatch causes incomplete noise cancellation
Solution Approach 1:
The patent applies local quality by creating a matching array with pixels that have identical physical characteristics and parasitic capacitance values to the imaging pixels. The matching array is positioned adjacent to the imaging array and shares the same pixel structure, ensuring that local parasitic effects are matched while using the reference array to suppress row-wise noise.
Solution Approach 2:
The patent creates a copy of the imaging array structure in the form of a matching array that replicates the pixel layout, dimensions, and electrical characteristics. This copy is used to generate reference signals that accurately represent the parasitic capacitance effects of the imaging array, enabling complete noise cancellation when the reference signals are subtracted from the imaging signals.
3Measurement precision
If multiple rows of non-imaging array are used, then noise sampling is improved, but offsets and variations in non-imaging rows create artifacts
Solution Approach 1:
The patent extracts only the essential reference information from a single row of the reference array that has been verified to be free of offsets and variations. By selecting one clean reference row and using it exclusively for noise cancellation, the system avoids incorporating artifacts from other reference rows that may contain offsets or variations, thus preventing artifact propagation to the output image.
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 effectively suppresses row-wise noise by using a global reference row for each imaging row, eliminating variations and parasitic capacitance mismatches, resulting in improved image quality with reduced artifacts and enhanced noise cancellation.
Implementation Method 1
The pixels are photodiodes that output a signal in response to incident light
Data Source
AI summary
An image sensor device is provided that has an uncovered imaging array of pixels and a covered global reference non-imaging array of pixels. The pixel samples of the global reference non-imaging array are used to remove noise from the pixel samples of the imaging array. The control signals and control lines that are used to sample the pixels of the imaging array are separate from and independent of the control signals and control lines that are used to sample the pixels of the global reference non-imaging array of pixels. For each row of pixels of the imaging array that is sampled, the same row of pixels of the global reference non-imaging array is sampled. The global reference row has no or very few offsets or variations to ensure that noise removal is performed effectively.