Image Sensor Segmentation for Motion Blur and Noise Trade-off

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Capturing quality images in low light or fast motion scenarios is challenging due to trade-offs between motion blur and noise in existing image sensor technologies, where long exposures reduce noise but introduce motion blur, and short exposures reduce blur but increase noise.

Innovation Solution

The method involves capturing multiple images with overlapping and non-overlapping exposure periods using different fields of pixels in an image sensor, allowing for spatial and temporal alignment to produce images with reduced motion blur and noise, by transferring charge packets to VCCDs and storing them for later readout.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a long exposure is used to increase the amount of light collected, then noise in the image is reduced, but motion blur artifacts appear in the image

Engineering Contradiction:

Improveimage qualityVSAvoidmotion blur

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The image sensor is divided into multiple independent fields (first field and second field) that can capture images simultaneously with different exposure times. This segmentation allows the system to capture both long exposure images (for low noise) and short exposure images (for reduced motion blur) at the same time, resolving the contradiction between noise reduction and motion blur prevention.

Inventive Principle:

Principle #1Segmentation

2Object-affected harmful factors

If a short exposure is used to reduce motion blur, then motion blur is reduced, but noise increases due to low signal at the image sensor

Engineering Contradiction:

Improvemotion blurVSAvoidimage quality

Core Design Contradiction:

Object-affected harmful factorsVSReliability

Solution Approach 1:

Inventive Principle:

Principle #1Segmentation

3Object-affected harmful factors

If multiple short exposure images are captured in rapid succession, then motion blur is reduced, but the time between captures increases due to readout requirements

Engineering Contradiction:

Improvemotion blurVSAvoidtime between captures

Core Design Contradiction:

Object-affected harmful factorsVSLoss of time

Solution Approach 1:

The image sensor is divided into multiple independent fields that can be read out simultaneously through separate charge-coupled devices. This parallel readout architecture eliminates the sequential readout bottleneck, allowing multiple images to be captured and read out without increasing the time between captures, thus maintaining rapid succession capability.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a spatial dimension to the readout process by using multiple charge-coupled devices (one for each field) that operate in parallel. This transforms the readout from a single-channel sequential process to a multi-channel parallel process, effectively reducing the time between captures while maintaining the ability to capture multiple images in rapid succession.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

4Productivity

If different rows of pixels are used for long and short exposure images, then two images can be captured, but the method is limited to only two images per set

Engineering Contradiction:

Improvenumber of images capturedVSAvoidflexibility in exposure combinations

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The image sensor is divided into multiple independent fields (first field, second field, and potentially additional fields) that can operate independently with different exposure times. This segmentation enables the capture of multiple images (not limited to two) with varying exposure durations, allowing flexible combinations of long and short exposure images to be captured simultaneously or in rapid succession.

Inventive Principle:

Principle #1Segmentation

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 enables the production of improved images with reduced motion blur and noise, allowing for accurate representation of scene motion and improved image quality in low light conditions.

Implementation Method 1

The photosensitive areas in all of the pixels are first reset as a group. The different fields of pixels are then effectively reset individually by transferring the charge packets into the Vertical Charge-Coupled Devices (VCCDs) or charge-to-voltage conversion regions.

Methodology Applied

Scientific EffectPhotoelectric effect: Photoelectric Effect

Data Source

PatentUS8144220B2Methods for capturing and reading out images from an image sensor

Publication Date: 2012.03.27 SEMICON COMPONENTS IND LLC

AI summary

Multiple images are captured where the exposure times for some of the images overlap and the images are spatially overlapped. Charge packets are transferred from one or more portions of pixels after particular integration periods, thereby enabling the portion or portions of pixels to begin another integration period while one or more other portions of pixels continue to integrate charge. Charge packets may be binned during readout of the images from the image sensor. Comparison of two or more images having different lengths of overlapping or non-overlapping exposure periods provides motion information. The multiple images can then be aligned to compensate for motion between the images and assembled into a combined image with an improved signal to noise ratio and reduced motion blur.