Flexible Computational Image Sensor for Motion-Blur-Free Compressive Sensing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image sensors face challenges in efficiently capturing high-speed, high dynamic range, and motion-blur free imaging due to long integration times, which result in motion blur and data processing complexities, especially in low light conditions, requiring innovative data compression and processing methods.

Innovation Solution

A sensor architecture with a photodetector array on one die layer and a register array on another, utilizing a pattern generator to create shutter patterns that control charge accumulation and readout, enabling efficient data compression and processing through temporal, spatial, or spatial-temporal compression, allowing for high-speed imaging and motion-blur free image formation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If shorter integration time and multiple frame capturing are used to reduce motion blur, then motion blur is reduced, but large amounts of data are generated that need efficient storage and transfer

Engineering Contradiction:
Improvemotion blur reductionVSAvoiddata volume
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The pixel array is divided into multiple independently controllable groups, where each group can be selectively activated or deactivated by individual shutter control signals. This segmentation allows the system to capture multiple frames with different spatial regions active, reducing the total data volume while maintaining motion blur reduction capabilities.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The shutter control signals are dynamically adjusted for each pixel group and frame, allowing adaptive control of which regions are captured at different times. This dynamic control enables the system to reduce data volume by only capturing relevant regions while maintaining the ability to reduce motion blur through multiple short-exposure frames.

Inventive Principle:
Principle #15Dynamics

2Illumination intensity

If multiple frames are captured with shorter integration times, then high dynamic range is achieved, but data processing complexity increases

Engineering Contradiction:
Improvehigh dynamic rangeVSAvoiddata processing complexity
Core Design Contradiction:
Illumination intensityVSDevice complexity

Solution Approach 1:

The pixel array is segmented into multiple groups that can be independently controlled, allowing different regions to capture different exposure levels simultaneously. This segmentation reduces the number of frames needed to achieve high dynamic range, thereby reducing processing complexity while maintaining HDR capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Instead of capturing complete frames for all regions at all exposure levels, the system applies partial action by selectively activating only the necessary pixel groups for each exposure level. This reduces the total number of frames captured and processed while still achieving the required dynamic range.

Inventive Principle:
Principle #16Partial or excessive action

3Quantity of substance

If traditional compressive sensing with optical masks is used, then data compression is achieved, but the masks need to be switched very quickly which is difficult

Engineering Contradiction:
Improvedata compressionVSAvoidmask switching speed
Core Design Contradiction:
Quantity of substanceVSSpeed

Solution Approach 1:

The patent replaces mechanical/optical mask systems with electrical shutter control signals that can be independently applied to different pixel groups. This substitution eliminates the need for physical mask switching, achieving the same compression effect through electronic control that can operate at much higher speeds.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The shutter control signals provide dynamic, electronically-controlled modulation of pixel group activation without requiring physical mask movement. This dynamic control achieves compression ratios comparable to optical masks but at significantly higher switching speeds limited only by the electronic control circuitry.

Inventive Principle:
Principle #15Dynamics

4Productivity

If compressive sensing patterns are applied at high speeds, then high frame rate is achieved, but storing mask values for each pixel frame requires a lot of memory

Engineering Contradiction:
Improveframe rateVSAvoidmemory requirements
Core Design Contradiction:
ProductivityVSQuantity of substance

Solution Approach 1:

The pixel array is divided into multiple groups that share common control signals, reducing the total number of unique shutter patterns needed. This segmentation decreases memory requirements for storing control patterns while maintaining high frame rates through efficient reuse of control signals across multiple pixel groups.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The shutter control system is designed to be universal, where a single control signal can activate or deactivate multiple pixel groups simultaneously. This multi-functionality reduces the number of unique control patterns needed in memory, achieving high frame rates with reduced memory requirements.

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

This approach enables efficient data compression and processing, reducing motion blur and enhancing image quality in high-speed and high dynamic range imaging, while allowing for compact sensor design with improved sensitivity and frame rate.

Implementation Method 1

a photodetector array located on a first die layer

Methodology Applied
Scientific EffectPhotoelectric effect: Photoelectric Effect

Data Source

PatentEP4207792A1A flexible computational image sensor with compressive sensing capability
Publication Date: 2023.07.05 SAMSUNG ELECTRONICS CO LTD
  • EP4207792A1 patent drawingFigure 1
  • EP4207792A1 patent drawingFigure 2
  • EP4207792A1 patent drawingFigure 3

AI summary

The present disclosure relates to a method and system for imaging a scene. The method includes generating a shutter pattern and applying the shutter pattern to a photodetector array. The system includes a sensor architecture in three dimensions, where elements of the sensor architecture are stacked in two or more layers. Some elements of the sensor architecture include a photodetector array, register array, a generator to generate shutter patterns, readout circuitry, and an ISP.