Image Sensor Noise Correction Using Pre-Capture RTS Data

Solution Overview

Problem

Miniaturization of image sensors leads to reduced sensitivity and increased noise, particularly blinking defect noise due to random telegraph signal noise caused by reading circuits, which existing technologies struggle to effectively correct.

Innovation Solution

An image processing apparatus and method that acquires image data and noise information to correct pixel values in image data generated by an image sensor, using positional information of reading circuits and pixels to identify and correct blinking defect noise through time-direction and spatial-direction noise reduction processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a sharing pixel structure is adopted to improve sensitivity, then the aperture ratio increases and sensitivity improves, but blinking defect noise increases due to reading circuit noise

Engineering Contradiction:

ImprovesensitivityVSAvoidblinking defect noise

Core Design Contradiction:

ReliabilityVSObject-generated harmful factors

Solution Approach 1:

The patent performs preliminary noise level detection for each pixel before image capture, storing the detected RTS noise levels in advance. This allows the correction unit to use pre-acquired reference image data and previously detected noise levels to correct blinking defect noise in real-time during image processing, rather than detecting noise during the actual capture process.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent implements a feedback mechanism where the correction unit continuously compares current pixel values with reference image data from different time points, using the stored RTS noise levels to dynamically adjust corrections. The system feeds back corrected image data to improve subsequent corrections, adapting to changing noise conditions while maintaining the benefits of the sharing pixel structure.

Inventive Principle:

Principle #23Feedback

2Object-generated harmful factors

If existing RTS noise correction technology is used, then some blinking defect noise is corrected, but the correction is insufficient and complex processing is required

Engineering Contradiction:

Improveblinking defect noise correctionVSAvoidcorrection processing complexity

Core Design Contradiction:

Object-generated harmful factorsVSDevice complexity

Solution Approach 1:

The patent performs preliminary noise level detection for each pixel before image capture, storing the detected RTS noise levels in advance. This allows the correction unit to use pre-acquired reference image data and previously detected noise levels to correct blinking defect noise in real-time during image processing, rather than detecting noise during the actual capture process.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The correction unit uses each pixel's own historical reference image data and its own detected RTS noise level to perform self-correction. The system does not require complex external intervention or sophisticated algorithms, as each pixel corrects its own blinking defect noise using its stored reference data and noise characteristics, simplifying the overall processing architecture.

Inventive Principle:

Principle #25Self-service

3Area of stationary object

If miniaturization of pixels and reading circuit is implemented, then the area required for circuit is reduced and aperture ratio increases, but sensitivity reduction and noise increase occur

Engineering Contradiction:

Improvecircuit areaVSAvoidsensitivity

Core Design Contradiction:

Area of stationary objectVSReliability

Solution Approach 1:

The patent makes the reading circuit serve multiple functions: it not only reads pixel signals but also performs noise level detection and stores RTS noise characteristics for each pixel. The same reading circuit infrastructure is used for both signal reading and noise characterization, eliminating the need for separate dedicated noise detection circuits and maintaining area efficiency while improving sensitivity correction capability.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The correction unit uses each pixel's own historical reference image data and its own detected RTS noise level to perform self-correction. The system does not require complex external intervention or sophisticated algorithms, as each pixel corrects its own blinking defect noise using its stored reference data and noise characteristics, simplifying the overall processing architecture.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS10368017B2Image processing apparatus, image processing method, and computer-readable recording medium

Publication Date: 2019.07.30 OLYMPUS CORPORATION(JP)

AI summary

An image processing apparatus includes: an acquisition unit configured to acquire image data and noise information including one of positional information on a reading circuit in which blinking defect noise caused by the reading circuit occurs and positional information on each of pixels; and a correction unit configured to set the image data acquired by the acquisition unit as correction target image data, and correct a pixel value of a pixel of interest of the correction target image data based on the noise information and reference image data that is based on image data acquired at a time different from the correction target image data.