Image Sensor Signal Processing for kTC Noise Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image pickup elements, such as those using CCDs or CMOSs, face challenges in suppressing kTC noise, especially when full depletion is not achieved, leading to deterioration in image quality due to residual noise affecting RN noise.
Innovation Solution
A signal processing apparatus and method that performs A/D conversion of signals from unit pixels under different reset signal states and uses correlated double sampling to generate output signals, ensuring full depletion is prevented during photoelectric conversion, thereby effectively suppressing kTC noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full depletion is not performed in the pixel structure, then the photoelectric conversion can be maintained, but the kTC noise becomes larger and is more difficult to suppress
Solution Approach 1:
The patent segments the signal processing into multiple distinct phases: reset phase (with reset signal high), read phase (with reset signal low), and correlated double sampling phase. By separating the noise measurement and signal measurement into distinct segments, the kTC noise can be identified and removed through subtraction in the correlated double sampling process, allowing full depletion to be prevented while still achieving noise suppression.
Solution Approach 2:
The patent performs a preliminary reset operation before the actual photoelectric conversion and readout. During this reset phase, the reset signal is set high to reset the floating diffusion, and the resulting signal (containing kTC noise) is captured and stored. This preliminary action allows the noise component to be measured and subsequently removed from the final image signal through correlated double sampling.
2Object-affected harmful factors
If conventional kTC noise suppression methods are used, then some noise reduction is achieved, but residual kTC noise remains and adversely affects RN noise
Solution Approach 1:
The patent implements a feedback mechanism through correlated double sampling where the reset signal level is measured and fed back into the signal processing chain. The measured reset signal (containing kTC noise) is subtracted from the photoelectric conversion signal, creating a feedback loop that actively removes the noise component. This feedback approach ensures complete removal of kTC noise without leaving residual effects that would degrade image quality.
Solution Approach 2:
The patent introduces the reset signal as an intermediary element that carries the kTC noise information. By using this intermediary signal as a reference in the correlated double sampling process, the actual kTC noise present during photoelectric conversion can be identified and removed. The reset signal acts as a mediator between the noise source and the noise removal mechanism, enabling precise noise suppression.
3Object-affected harmful factors
If multiple A/D conversions and correlated double sampling are performed, then kTC noise is suppressed, but the processing complexity increases
Solution Approach 1:
The patent merges multiple signal processing operations into a unified correlated double sampling framework. The reset phase signal acquisition, photoelectric conversion signal acquisition, and noise removal operations are combined into a single integrated process flow. By merging these operations, the patent achieves effective kTC noise suppression while minimizing the increase in processing complexity that would result from completely separate processing chains.
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 approach significantly reduces kTC noise, thereby improving image quality by adapting the A/D conversion and correlated double sampling processes to manage noise effectively across different operational states of the unit pixels.
Implementation Method 1
unit pixels configured to perform photoelectric conversion of incident light
Implementation Method 2
reset noise to be generated, for example, by thermal fluctuation of the electric charge, that is, kTC noise is generated
Data Source
AI summary
Provided is a signal processing apparatus, including: an A/D conversion unit configured to perform A/D conversion of a first signal, A/D conversion of a second signal, A/D conversion of a third signal, and A/D conversion of a fourth signal; and a correlated double sampling processing unit configured to generate a first output signal by performing correlated double sampling using a first digital data item obtained through the A/D conversion of the first signal, and a second digital data item obtained through the A/D conversion of the second signal, a second output signal by performing correlated double sampling using a third digital data item obtained through the A/D conversion of the third signal, and a fourth digital data item obtained through the A/D conversion of the fourth signal, and a third output signal by performing correlated double sampling using the first output signal and the second output signal.