Image Sensor Noise Reduction via Asynchronous Reset Termination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image sensors face significant noise challenges, particularly reset noise (kT/C noise), which dominates in low light conditions and limits signal-to-noise ratio, as existing technologies struggle to effectively minimize thermal noise during the reset process.
Innovation Solution
The implementation of noise reduction circuitry that monitors pixel columns and terminates the reset sequence asynchronously when a targeted reference level is reached, utilizing existing control lines to reduce thermal noise by applying voltages that exceed the bandwidth of thermal noise, thereby reducing the sampled noise below the sqrt(kT/C) limit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a standard reset sequence is used to reset the pixel between frames, then the pixel is prepared for the next integration period, but thermal noise (kT/C noise) is introduced that dominates in low light conditions
Solution Approach 1:
The patent applies preliminary action by performing a first reset before the integration period to establish a known initial voltage state, then performing a second reset after integration to prepare for readout. This preliminary resetting action ensures that thermal noise is minimized at critical points while maintaining proper pixel operation throughout the frame cycle.
Solution Approach 2:
The patent implements feedback through correlated double sampling, where the voltage at the sense node is sampled twice (once before integration and once after), and the difference between these samples is taken. This feedback mechanism allows the system to measure and subtract thermal noise components from the final signal, significantly improving signal-to-noise ratio in low light conditions.
2Object-generated harmful factors
If the reset capacitance is increased to reduce thermal noise, then noise is reduced, but the pixel area and device complexity increase
Solution Approach 1:
Instead of increasing capacitance to reduce thermal noise, the patent uses feedback through correlated double sampling to measure and subtract noise components. This allows noise reduction without requiring larger capacitance values, thereby maintaining small pixel area while achieving superior signal-to-noise ratio.
Solution Approach 2:
The patent changes the temporal parameters of the reset operation by performing resets at specific times (before and after integration) rather than continuously. This time-based parameter control allows noise reduction through sampling strategy rather than through increasing hardware capacitance, avoiding pixel area expansion.
3Object-generated harmful factors
If additional circuitry is added to reduce thermal noise, then noise reduction is achieved, but device complexity and manufacturing cost increase
Solution Approach 1:
The patent achieves noise reduction using the existing sense node and readout circuitry that are already present in standard pixel designs. The same circuit elements perform multiple functions: integrating light signals, storing charge, and enabling correlated double sampling. This multi-functionality approach reduces thermal noise without adding dedicated noise reduction circuitry, keeping device complexity low.
Solution Approach 2:
The pixel circuit serves itself by using its own sense node to sample and compare voltages during the correlated double sampling process. The existing circuitry automatically performs noise measurement and subtraction without requiring external or additional specialized components, achieving noise reduction through self-service operation.
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 reduces thermal noise in image sensors, improving the signal-to-noise ratio and enhancing image quality, especially in low light conditions, without requiring additional circuitry or transistors, thus addressing the limitations of existing noise reduction methods.
Implementation Method 1
Thermal noise is electronic noise generated by the thermal agitation of charge carriers, such as electrons, in a circuit. The root-mean-square noise voltage, v, generated in a circuit is: v=sqrt(kT/C)
Data Source
AI summary
Systems and methods are disclosed for reducing reset noise in an image sensor. Voltage on the column read out is sensed during reset. When the voltage reaches a desired threshold level, a voltage is asserted on the column read out line that turns off the reset transistor. Using column circuitry to turn off the reset transistor may be used to reduce noise associated with the reset switch. In example embodiments, a comparator may be included on each column line to determine when the threshold voltage has been reached and to trigger the assertion of the turn off voltage on the column line.