Pixel-Calibrated Photon Counting for Low-Noise CMOS Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
CMOS image sensors face challenges in accurately counting photons due to readout noise and variations in gain and offset values among pixels, leading to potential misidentification of photon counts and degraded accuracy.
Innovation Solution
A photon counting device and method that utilize a conversion unit referencing gain and offset data for each pixel to accurately convert digital values to photon counts, using threshold values and correction parameters to mitigate the effects of pixel variations, with readout noise controlled at or below 0.2 e-rms to minimize incorrect detection rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If photon counting is performed using a CMOS image sensor with standard readout, then the device complexity is low and ease of operation is maintained, but readout noise and variations in gain and offset values cause degradation of photon counting accuracy
Solution Approach 1:
The patent applies preliminary action by measuring and storing the gain and offset values for each pixel before actual photon counting measurements. This pre-characterization allows the system to compensate for pixel variations during measurement by using the stored reference values to normalize the raw pixel signals, thereby improving photon counting accuracy without adding complexity to the measurement process itself
Solution Approach 2:
The patent applies parameter changes by transforming the raw pixel signal parameters (which are affected by readout noise and variations) into corrected photon count values through mathematical operations. By changing the parameter representation from raw digital values to normalized photon counts using stored gain and offset parameters, the system eliminates the harmful effects of pixel variations
2Measurement precision
If gain and offset variations among pixels are not compensated, then the device complexity remains low, but digital values for different photon counts may overlap leading to misidentification
Solution Approach 1:
The patent applies copying by creating a reference copy of each pixel's gain and offset characteristics during a calibration phase. These copied parameter values are stored and later used to correct the raw measurement signals, allowing accurate photon counting without requiring complex real-time adjustment mechanisms for each pixel
Solution Approach 2:
The patent introduces an intermediary processing step where stored gain and offset values act as mediators between the raw pixel signal and the final photon count determination. This intermediary correction layer simplifies the overall system by using pre-measured parameters to bridge the gap between variable pixel responses and standardized photon counting results
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 curbs the degradation of photon counting accuracy by ensuring accurate conversion of digital values to photon counts, even with variations in gain and offset values, achieving an incorrect detection rate of 1% or less with readout noise at or below 0.2 e-rms and 0.1% with noise at or below 0.15 e-rms.
Implementation Method 1
a photoelectric conversion element configured to convert input light to charge
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A photon counting device includes a plurality of pixels each including a photoelectric conversion element configured to convert input light to charge, and an amplifier configured to amplify the charge converted by the photoelectric conversion element and convert the charge to a voltage, an A/D converter configured to convert the voltages output from the amplifiers of the plurality of pixels to digital values; and a conversion unit configured to convert the digital value output from the A/D converter to the number of photons by referring to reference data, for each of the plurality of pixels, and the reference data is created based on a gain and an offset value for each of the plurality of pixels.