SPAD Image Sensor Photon Counting for Low Light Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image sensors, such as CMOS and CCD, suffer from high read noise and motion blur, especially under low light conditions and when used in head-mounted displays (HMDs) that undergo motion, affecting pass-through imaging and other operations like reprojection and object tracking.
Innovation Solution
The use of single photon avalanche diodes (SPADs) with sequential exposure and readout operations, where each SPAD pixel detects photons and generates binary outputs, allowing for high-framerate imaging with reduced noise and motion artifacts by binarizing the signal and combining partial frames to form a composite image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional CMOS or CCD image sensors are used under low light conditions, then the device can capture images, but the read noise dominates the signal and the signal-to-noise ratio deteriorates
Solution Approach 1:
The patent changes the operating parameters of the image sensor by using SPAD pixels that operate in photon-counting mode with binary output (0 or 1), fundamentally altering how light detection is performed. This parameter change enables the system to achieve high signal-to-noise ratio in low light by counting individual photons rather than measuring continuous analog signals that are susceptible to read noise
Solution Approach 2:
The patent replaces the conventional analog photoelectric conversion mechanism in CMOS/CCD sensors with a digital photon-counting mechanism using SPADs. This substitution eliminates the read noise problem inherent in analog systems by directly converting photon detection events into discrete digital counts, thereby improving measurement precision in low light conditions
2Measurement precision
If the frame rate is reduced to allow conventional sensors to detect enough photons, then the signal-to-noise ratio improves, but motion blur increases
Solution Approach 1:
The patent enables continuous high-rate photon detection by using SPAD pixels that can rapidly reset and count photons at high frame rates (e.g., 90 Hz or higher). The binary nature of SPAD output allows for continuous accumulation of photon counts across multiple frames without the motion blur penalty that would require lowering the frame rate in conventional sensors
Solution Approach 2:
The patent employs periodic shutter operations to control exposure timing for each frame while maintaining high overall frame rates. By using short, periodic exposure windows with SPAD pixels that accumulate photon counts over multiple periods, the system achieves both high temporal resolution (reducing motion blur) and sufficient photon detection (maintaining signal-to-noise ratio)
3Speed
If high frame rate imaging is performed with conventional sensors, then motion blur is reduced, but read noise dominates the signal under low light conditions
Solution Approach 1:
The patent replaces the analog signal integration mechanism in conventional sensors with a digital photon-counting mechanism using SPADs. This substitution allows the system to maintain high frame rates while improving signal-to-noise ratio, as each SPAD pixel independently counts photons without being affected by read noise that plagues analog systems at high sampling rates
Solution Approach 2:
The patent applies preliminary shutter operations to each SPAD pixel before photon detection begins, configuring the pixels to be ready for high-rate counting. This preliminary preparation enables the system to immediately start high frame rate imaging without losing photons during sensor initialization, thereby maintaining both high speed and measurement precision from the first frame
4Measurement precision
If sequential exposure and readout operations are performed with SPAD pixels, then noise is reduced and signal-to-noise ratio improves, but the device complexity increases
Solution Approach 1:
The patent divides the image capture process into multiple sequential exposure and readout operations, with each SPAD pixel independently performing photon counting during each exposure period. This segmentation allows for sophisticated noise reduction through temporal accumulation of binary counts while keeping each individual pixel operation simple and identical, thereby improving measurement precision without proportionally increasing device complexity
Solution Approach 2:
The patent enables each SPAD pixel to autonomously perform photon detection and binary output generation without requiring complex readout circuitry per pixel. The simplicity of the binary output (0 or 1) from each pixel allows for efficient parallel processing and accumulation across multiple frames, achieving high signal-to-noise ratio through the self-service capability of individual pixels rather than through complex centralized processing
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 significantly improves the signal-to-noise ratio and reduces motion blur, enabling high-framerate low-light imaging with reduced noise, even in HMDs, and enhances operations like reprojection and object tracking.
Implementation Method 1
each pixel is configured to generate electron-hole pairs in response to detected photons
Implementation Method 2
single photon avalanche diodes (SPADs)
Implementation Method 3
generate electron-hole pairs in response to detected photons
Data Source
AI summary
A system for image acquisition with reduced noise using SPADs is configured to perform a plurality of sequential exposure and readout operations. Each exposure and readout operation includes (i) applying a set of shutter operations to configure each SPAD pixel of the SPAD array to enable photon detection, and (ii) for each SPAD pixel of the SPAD array, reading out a number of photons detected during the set of shutter operations. The system is also configured to generate an image based on the number of photons detected for each SPAD pixel during each of the plurality of sequential exposure and readout operations.