Image Sensor Noise Reduction via Temporal Frame Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image sensors face challenges in reducing photon shot noise, particularly in smaller pixel sizes, as increasing pixel capacitance or voltage swing is not feasible due to size constraints, and existing noise reduction techniques are not effective for high-quality images in average or good lighting systems.
Innovation Solution
The image sensor operates with a pixel readout circuit that produces first image frames at a faster rate than the image output circuit, allowing for statistical combination of multiple frames to produce second image frames, with an exposure monitoring circuit adjusting the frame rate based on pixel values to optimize noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If pixel size is reduced to decrease device size, then device size is reduced, but full well capacity is reduced and noise increases
Solution Approach 1:
The patent combines multiple frames captured at different times into a single output frame by averaging their pixel values. This merging process effectively increases the signal-to-noise ratio, compensating for the reduced full well capacity of smaller pixels and thereby reducing noise in the final image.
Solution Approach 2:
The patent transitions from spatial noise reduction (combining adjacent pixels) to temporal noise reduction (combining pixels across time). By capturing and averaging multiple frames over time, the system reduces noise without requiring larger pixel areas, thus maintaining small device size while improving measurement precision.
2Measurement precision
If full well capacity is increased to reduce photon shot noise, then photon shot noise is reduced, but pixel size must be increased
Solution Approach 1:
The patent performs preliminary capture of multiple frames before final image output. By capturing several frames in advance and averaging them, the system effectively increases the integrated signal without requiring larger pixels, thereby reducing photon shot noise while maintaining small pixel dimensions.
Solution Approach 2:
The patent maintains continuous capture of frames at a high frame rate, utilizing the continuous accumulation of signal over time. This continuous useful action allows the system to achieve effective full well capacity enhancement through temporal integration rather than spatial expansion.
3Speed
If frame rate is increased to reduce motion artifacts, then motion artifacts are reduced, but exposure time per frame is reduced and noise increases
Solution Approach 1:
The patent merges multiple high-frame-rate captures by averaging pixel values across frames. This combining process maintains the benefits of high frame rate (reduced motion artifacts) while compensating for the reduced exposure time per frame through temporal signal accumulation, thereby reducing noise in the final image.
Data Source
AI summary
An imaging system includes a plurality of pixels. A pixel readout circuit produces a plurality of first image frames from those pixels. An image output circuit produces a plurality of second image frames and operates to produce a second image frame from more than one of the first image frames. The pixel readout circuit is enabled to produce the first images frames at a rate faster than the image output circuit produces the second image frames. Through combining first image frames, by averaging or other statistical combinations, the photon shot noise of second image frames is reduced. Photon shot noise affects images with high light levels more than those with low light levels and, as such, the system processing alters the rate of first image frames dependent on the current light levels.