Imaging Device Crosstalk Removal via Pixel Block Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing imaging devices that acquire multispectral images using a microlens array suffer from light leaks and crosstalk issues, requiring complex calculations to determine and correct for varying amounts of crosstalk across pixels, which complicates the acquisition of high-quality images.

Innovation Solution

An imaging device with a configuration of pixel blocks incorporating first and second optical filter elements, polarization filter units, and a storage unit that uses a matrix to calculate pixel signals and remove crosstalk, allowing for the acquisition of high-quality multispectral images by processing signals from each pixel block to isolate signals corresponding to each optical region.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a microlens array is used to separate beams from different optical regions, then multispectral images can be acquired with one image sensor, but light leaks to adjacent pixels and crosstalk occurs

Engineering Contradiction:
Improveimage acquisition efficiencyVSAvoidsignal accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The pixel array is divided into multiple pixel blocks, where each pixel block contains multiple pixels with different optical filter combinations. This segmentation allows each pixel block to independently receive and process light from different optical regions, reducing crosstalk between adjacent pixels while maintaining efficient multispectral image acquisition.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different optical filter elements are selectively arranged in different pixels within each pixel block. Specifically, pixels are configured with different combinations of first optical filters (with different spectral transmittances) and second optical filters (with different polarization directions), creating local quality variations that enable each pixel to respond differently to light from specific optical regions, thereby reducing crosstalk.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If signal processing is performed to remove crosstalk influence, then image quality improves, but calculation complexity increases due to varying crosstalk amounts across pixels

Engineering Contradiction:
Improveimage qualityVSAvoidsignal processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent pre-calculates and stores crosstalk correction coefficients in a storage unit during the device manufacturing or initialization phase. These coefficients are derived from the known optical characteristics and geometric relationships of the system. During actual image acquisition, the processor simply retrieves and applies these pre-computed coefficients to correct crosstalk, avoiding complex real-time calculations while maintaining high image quality.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If multiple optical filter elements are combined in each pixel block, then crosstalk is reduced and signal accuracy improves, but device structure becomes more complex

Engineering Contradiction:
Improvesignal accuracyVSAvoidpixel structure complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

Each pixel block serves multiple functions: it captures light from multiple optical regions, performs spectral filtering, and executes polarization filtering, all within a compact structure. The combination of first optical filters (spectral transmittance) and second optical filters (polarization direction) in each pixel block enables multi-functional operation, reducing the need for separate dedicated components for each function.

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

Solution Approach 2:

The patent merges spectral filtering and polarization filtering functions into a single integrated pixel block structure. Multiple optical filter elements are combined within each pixel block, allowing simultaneous execution of spectral and polarization filtering operations in one compact unit, thereby reducing overall device complexity while maintaining high signal accuracy.

Inventive Principle:
Principle #5Merging (Combining)

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 enables the acquisition of high-quality multispectral images by effectively removing crosstalk and ensuring uniform light incidence on each pixel block, simplifying the image generation process and improving image quality.

Implementation Method 1

n types of first optical filter elements having different spectral transmittances

Methodology Applied
Scientific EffectSpectral transmittance: Filter (optical)

Implementation Method 2

m types of second optical filter elements having different transmission polarization directions

Methodology Applied
Scientific EffectPolarization: Polarisation

Implementation Method 3

a microlens array is disposed in front of the image sensor. The imaging device acquires a multispectral image with one image sensor by guiding light from different optical regions of the pupil portion to different pixels using the beam separation effect of the microlens array.

Methodology Applied
Scientific EffectBeam separation: Lens

Data Source

PatentEP3866462B1Imaging device
Publication Date: 2022.11.02 FUJIFILM CORP
  • EP3866462B1 patent drawingFigure 1
  • EP3866462B1 patent drawingFigure 2~3
  • EP3866462B1 patent drawingFigure 4

AI summary

Provided is an imaging device that can acquire a high-quality multispectral image with one image sensor. An imaging device includes: an image sensor (100) including a plurality of pixel blocks each of which includes n (n ≥ 2) types of spectral filter elements having different spectral transmittances and m (2 ≤ m ≤ 3) types of polarization filter elements having different transmission polarization directions which are combined such that q (q = n × m) types of light are received by each pixel; an optical system (10) that includes k (k ≤ q) optical regions transmitting light in different wavelength bands and polarization filter units provided in each of the optical regions; and a signal processing unit (200) that processes a signal (pixel signal) of each pixel obtained from the image sensor (100) to generate an image of each optical region of the optical system (10). The signal processing unit (200) performs predetermined arithmetic processing on q pixel signals (x1, x2, ..., Xq) obtained from each pixel block of the image sensor (100) to calculate k pixel signals (X1, X2, ..., Xk) corresponding to each of the optical regions of the optical system (100) and generates an image of each optical region.