CMOS Image Sensor Dynamic Range Extension via Pixel Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
CMOS image sensors in mobile applications face limitations in dynamic range extension due to constraints in pixel size and chip size, leading to inadequate reproduction of high and low light conditions, with existing techniques such as multiple integration, logarithmic sensors, and spatial-varying sensitivity pixels being inadequate.
Innovation Solution
The solution involves setting different integration times and signal gains for each pixel in a pixel array, grouping corresponding digital data from adjacent pixels into superpixels with unique combinations of integration time and signal gain, and using column analog-to-digital converters to generate digital data, allowing for dynamic adjustment based on scene brightness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple integration approach is used to extend dynamic range, then dynamic range is improved, but extra memory is required which increases device complexity
Solution Approach 1:
The pixel array is divided into multiple regions, each with different integration times. This segmentation allows different parts of the sensor to capture light for different durations simultaneously, extending dynamic range without requiring extra memory to store multiple full images.
Solution Approach 2:
The integration time is made dynamic and spatially varying across the pixel array. Different regions can have different integration times adapted to local lighting conditions, allowing the sensor to optimize dynamic range capture without fixed hardware additions.
2Measurement precision
If logarithmic sensor approach is used to extend dynamic range, then dynamic range is improved, but pixel response non-uniformity increases which degrades color response
Solution Approach 1:
Different regions of the pixel array are assigned different integration times based on local lighting conditions. This local adaptation allows each region to optimize its exposure independently, maintaining pixel response uniformity within each region while achieving extended dynamic range across the entire image.
3Measurement precision
If dual sensitivity pixels approach is used to extend dynamic range, then dynamic range is improved, but extra transistors and storage inside pixel are required which prevents pixel size reduction
Solution Approach 1:
Instead of adding complexity within each pixel (extra transistors and storage), the solution moves to a spatial dimension by creating multiple regions across the pixel array with different integration times. This dimensional shift achieves extended dynamic range without increasing individual pixel size.
4Measurement precision
If spatial-varying sensitivity pixels with optical masks are used to extend dynamic range, then dynamic range is improved, but pixel response becomes nonlinear requiring full calibration and lookup tables
Solution Approach 1:
The integration time parameter is varied across different spatial regions of the pixel array. By changing this temporal parameter rather than using optical masks that alter spatial sensitivity, the system achieves extended dynamic range with linear pixel responses that do not require complex calibration or lookup tables.
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 extends the dynamic range of CMOS image sensors, enabling the capture of high and low light conditions without increasing pixel size or chip complexity, resulting in improved image quality with reduced spatial resolution.
Implementation Method 1
light is passed through the camera lens to be exposed on an array of pixels having photodiodes such that the captured amount of light in each pixel can be converted into analog electric signals
Data Source
AI summary
Aspects of the invention provide dynamic range extension for CMOS image sensors for mobile applications. An embodiment of the invention may comprise setting for each pixel in a pixel array one of a plurality of integration times and one of a plurality of signal gains, wherein the settings may be used to generate corresponding digital data for each pixel in the pixel array. The corresponding digital data for adjacent pixels for the same color plane may then be grouped into a superpixel, where each pixel has associated with it a different combination of integration time and signal gain.