Imaging Device Multi-Gain Amplifier Noise Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Imaging devices struggle to reduce noise in synthesized images, especially in low-luminance portions, when using multiple gains for amplification, leading to conspicuous random noise and difficulty in determining the black level of pixel signals.

Innovation Solution

An imaging device with an amplifier capable of multiple gain settings and a signal processing circuit that reads out noise and pixel signals with different gains, where the first gain is larger than the second gain, allowing for effective noise reduction and dynamic range expansion.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Illumination intensity

If multiple gains are used for amplification in synthesized images, then dynamic range is expanded, but random noise becomes conspicuous in low-luminance portions

Engineering Contradiction:
Improvedynamic rangeVSAvoidrandom noise
Core Design Contradiction:
Illumination intensityVSObject-affected harmful factors

Solution Approach 1:

The pixel signal is segmented into multiple components through correlated double sampling: a first pixel signal component is extracted using a first gain, and a second pixel signal component is extracted using a second gain. This segmentation allows selective use of appropriate gain levels for different signal conditions, reducing noise in low-luminance portions while maintaining dynamic range expansion capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The amplification gain parameter is dynamically changed based on signal conditions. The system switches between first gain and second gain values depending on the luminance level and signal characteristics, optimizing the signal-to-noise ratio while preserving the expanded dynamic range for both bright and dark regions.

Inventive Principle:
Principle #35Parameter changes

2Illumination intensity

If multiple gains are used for amplification, then dynamic range is expanded, but determination of black level becomes difficult

Engineering Contradiction:
Improvedynamic rangeVSAvoidblack level determination
Core Design Contradiction:
Illumination intensityVSMeasurement precision

Solution Approach 1:

A black level signal is preliminarily determined before actual image acquisition by setting all pixel electrodes to a reference potential and measuring the output signal. This preliminary black level determination is performed separately for each gain value, establishing accurate reference points that enable precise black level correction in the final synthesized image regardless of which gain was used.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from the correlated double sampling process to automatically determine and correct the black level. By comparing the first and second pixel signal components and analyzing their differences, the system feeds back adjustment information to accurately establish the black level position, resolving the ambiguity caused by multiple gain values.

Inventive Principle:
Principle #23Feedback

3Object-affected harmful factors

If correlated double sampling is performed with large gain, then noise suppression effect is large, but device complexity increases

Engineering Contradiction:
Improvenoise suppressionVSAvoidcircuit complexity
Core Design Contradiction:
Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The same amplifier circuit performs multiple functions: it amplifies signals with different gain values and enables correlated double sampling operations. By making the amplifier multi-functional rather than requiring separate dedicated circuits for each gain level and sampling operation, the patent achieves strong noise suppression through correlated double sampling without proportionally increasing device complexity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

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 solution significantly reduces random noise in low-luminance areas and enables accurate determination of the black level, improving the dynamic range and noise characteristics of synthesized images.

Implementation Method 1

a pixel unit including a plurality of photoelectric converters

Methodology Applied
Scientific EffectPhotoelectric conversion: Photoelectric Effect

Data Source

PatentUS11165978B2Imaging device, control method thereof, and imaging apparatus
Publication Date: 2021.11.02 CANON KK
  • US11165978B2 patent drawing
  • US11165978B2 patent drawing
  • US11165978B2 patent drawing

AI summary

In order to provide an imaging device that includes an amplifier performing amplification with a plurality of gains for each signal from a pixel unit and can further reduce a noise component, an imaging device included in an imaging apparatus has the pixel unit in which unit pixels are arranged in a matrix and generates a signal voltage by photoelectric conversion. A column amplifier can amplify a photoelectrically converted signal with a plurality of gains. After an amplified signal is subjected to analog/digital conversion by a column ADC, a signal processing circuit subtracts a noise signal from a pixel signal. A noise signal read out after resetting a floating diffusion section included in the pixel unit and a pixel signal read out immediately thereafter are acquired by being multiplied by a first gain, and a pixel signal read out thereafter is acquired by being multiplied by a second gain. The first gain is set to a value larger than the second gain.