Constrained Cross-Talk Templates for SPAD Dark Current Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image sensors, such as CMOS and CCD, suffer from high read noise and motion blur under low light conditions, affecting pass-through imaging and other operations in mixed-reality systems, while single photon avalanche diode (SPAD) sensors face challenges with dark current noise, particularly under low light conditions.
Innovation Solution
The implementation of techniques to generate dark current residual images that compensate for dark current noise in SPAD imagery by using template matching, non-maximum suppression, and weighted filtering to refine dark current images, combined with motion compensation and ambient light scaling to improve image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If conventional image sensors (CMOS/CCD) are used for low light imaging, then the sensor can capture images in dark environments, but read noise and fixed pattern noise dominate the signal and decrease signal-to-noise ratio
Solution Approach 1:
The patent replaces conventional CMOS/CCD image sensors with single-photon avalanche diode (SPAD) sensors that utilize quantum mechanical effects (photon detection through avalanche multiplication) to achieve superior low-light performance. SPAD sensors can detect individual photons and operate effectively in extremely low light conditions where conventional sensors fail due to read noise domination.
2Speed
If conventional image sensors are used at high frame rates under low light conditions, then motion capture is improved, but read noise dominance and motion blur increase
Solution Approach 1:
The patent replaces conventional image sensors with SPAD sensors that utilize photon-time-of-flight measurement and histogram-based depth mapping to achieve high frame rates without motion blur. The SPAD array captures photon arrival times and constructs depth images through temporal histogram accumulation, enabling high-speed operation in low light without the read noise and motion blur problems of conventional sensors.
3Measurement precision
If dark current compensation techniques are applied to SPAD sensors, then dark current noise is reduced, but device complexity increases
Solution Approach 1:
The patent applies preliminary dark current compensation by capturing dark current images at the same temperature and exposure time as regular images, then subtracting these dark current images from the captured images to remove dark current noise. This preliminary compensation approach effectively reduces dark current noise without requiring complex real-time processing during normal operation.
Solution Approach 2:
The patent creates a copy of the dark current characteristics by capturing dark current images under identical conditions (temperature, exposure time) as the regular images. These dark current copies are then used to compensate and remove dark current noise from the captured images, providing an effective and simple compensation mechanism.
4Measurement precision
If factory calibrated dark current images are used for compensation, then dark current noise is reduced, but the compensation becomes inaccurate over time due to sensor degradation
Solution Approach 1:
The patent performs preliminary dark current compensation by capturing dark current images at the same temperature and exposure time as regular images, then subtracting these dark current images from the captured images to remove dark current noise. This preliminary compensation approach effectively reduces dark current noise without requiring complex real-time processing during normal operation.
Solution Approach 2:
The patent implements a self-updating dark current compensation mechanism where the system automatically captures dark current images and updates the dark current lookup table during operation. This self-service approach ensures that dark current compensation remains accurate over time by continuously adapting to sensor degradation and environmental changes without requiring external recalibration.
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
The proposed methods significantly reduce dark current noise and improve image quality in SPAD sensors, enhancing performance under low light conditions and mitigating artifacts caused by faulty pixels.
Implementation Method 1
A SPAD pixel is operated at a bias voltage that enables the SPAD to detect a single photon. Upon detecting a single photon, an electron-hole pair is formed, and the electron is accelerated across a high electric field, causing avalanche multiplication
Implementation Method 2
image sensing pixel arrays where each pixel is configured to generate electron-hole pairs in response to detected photons
Data Source
AI summary
A system for modifying a dark current image is configurable to receive an input image depicting a dark current state for one or more pixels of the input image, The dark current state for one or more pixels of the input image comprises one of: a faulty state or a non-faulty state. The system is configurable to partition the input image into a plurality of partitions and generate an updated input image by imposing at least one quantity constraint or at least one severity constraint to the plurality of partitions in association with at least one type of dark current state.