IP Library Granted Patent US 12,666,167
Granted Patent B2
US 12,666,167 · App. 18/602,814 · Granted Jun 23, 2026

System and method for multispectral extraction from crosstalk profile

Inventors: Rivka Emanuel (Tel-Aviv, IL); Gal Bitan (Tel-Aviv, IL); Elad Yaakobi (Tel-Aviv, IL); Elena Grechikhin (Tel-Aviv, IL); Shahaf Duenyas (Tel-Aviv, IL)
Assignee: SAMSUNG ELECTRONICS CO., LTD.
H04N25/134H04N23/85H04N25/62
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,666,167
App. No.
18/602,814
Filed
Mar 12, 2024
Granted
Jun 23, 2026
Kind
B2
Art Unit
2639
USPC
348/222.1
Abstract

An electronic device includes: memory; a color filter array (CFA); an image sensor operatively connected to the CFA, the image sensor being configured to capture a plurality of images respectively having a plurality of measured spectral densities in a plurality of spectral bands; and a processor operatively connected to the at least one memory and the image sensor, wherein the at least one processor is configured to: measure a plurality of crosstalk values in a plurality of pixels and in a plurality of K channels of the plurality of images (the K represents a number of pixels per a color of the CFA), estimate a plurality of spectral-crosstalk values by applying an estimation function to the plurality of crosstalk values, and control to store, in at least one of the memory, the image sensor, or the processor, the estimated plurality of spectral-crosstalk values.

Claims (41)

1 . An electronic device comprising:

at least one memory;

a color filter array (CFA);

an image sensor operatively connected to the CFA, the image sensor being configured to capture a plurality of M images respectively having a plurality of measured spectral densities in a plurality of D spectral bands; and

at least one processor operatively connected to the at least one memory and the image sensor,

wherein the at least one processor is configured to:

measure a plurality of crosstalk values in a plurality of L pixels and in a plurality of K channels of the plurality of M images, wherein the K represents a number of pixels per a color of the CFA,

estimate a plurality of spectral-crosstalk values by applying an estimation function to the plurality of crosstalk values, and

control to store, in at least one of the at least one memory, the image sensor, or the at least one processor, the estimated plurality of spectral-crosstalk values.

2 . The electronic device of claim 1 , wherein the CFA is an up-sampled Bayer CFA.

3 . The electronic device of claim 1 , wherein the CFA is Quad Bayer, Nona pixel, or Tetra pixel.

4 . The electronic device of claim 1 , wherein the estimation function corresponds to a pseudoinverse of a spectral matrix or a neural network model.

5 . The electronic device of claim 4 , wherein the plurality of measured spectral densities are arranged in the spectral matrix.

6 . The electronic device of claim 1 , wherein the at least one processor is further configured to measure the plurality of crosstalk values in the plurality of L pixels and in the plurality of K channels of the plurality of M images, based on a plurality of deviations between pixel levels of the plurality of L pixels and average color values of the plurality of K channels.

7 . The electronic device of claim 1 , wherein the at least one processor is further configured to capture the plurality of M images using an integrating light sphere configured to measure ground-truth spectral densities of the plurality of M images.

8 . An electronic device comprising:

at least one memory;

a color filter array (CFA);

an image sensor operatively connected to the CFA, the image sensor being configured to capture a plurality of M images respectively having a plurality of measured spectral densities in a plurality of D spectral bands; and

at least one processor operatively connected to the at least one memory and the image sensor,

wherein the at least one processor is configured to:

measure a plurality of crosstalk values in a plurality of L pixels and in a plurality of K channels of the plurality of M images, wherein the K represents a number of pixels per a color of the CFA,

estimate a plurality of spectral-crosstalk values by applying an estimation function to the plurality of crosstalk values, and

control to store, in at least one of the at least one memory, the image sensor, or the at least one processor, the estimated plurality of spectral-crosstalk values,

wherein the electronic device further comprises a flat detector operatively connected to the image sensor,

wherein the at least one processor is further configured to:

receive an input image having a pattern of the CFA,

receive, from the flat detector, info on a first area of the input image, wherein the first area has a first crosstalk value,

calculate the first crosstalk value in the first area of the input image,

estimate a first spectral density of the first area by applying the pseudoinverse of the plurality of estimated spectral-crosstalk values to the calculated first crosstalk value, and

control to store, in at least one of the at least one memory, the image sensor, or the at least one processor, the estimated first spectral density of the first area, and

wherein the CFA is an up-sampled Bayer CFA.

9 . The electronic device of claim 8 , wherein the at least one processor is further configured to:

receive, from the flat detector, info on a second area of the input image, wherein the second area has a second crosstalk value,

calculate the second crosstalk value in the second area of the input image,

estimate a second spectral density of the second area by applying a pseudoinverse of the plurality of estimated spectral-crosstalk values to the calculated first crosstalk value,

control to store, in at least one of the at least one memory, the image sensor, or the at least one processor, the estimated second spectral density of the second area, and

generate a global spectral density estimate of the input image by aggregating the estimated first spectral density of the first area and the estimated second spectral density of the second area.

10 . The electronic device of claim 8 , wherein the CFA is Quad Bayer, Nona pixel, or Tetra pixel.

11 . The electronic device of claim 8 , wherein the at least one processor is further configured to measure the plurality of crosstalk values in the plurality of L pixels and in the plurality of K channels of the plurality of M images, based on a plurality of deviations between pixel levels of the plurality of L pixels and average color values of the plurality of K channels.

12 . The electronic device of claim 8 , wherein the at least one processor is further configured to capture the plurality of M images using an integrating light sphere configured to measure ground-truth spectral densities of the plurality of M images.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Mar 13, 2024
From: EMANUEL, RIVKA; BITAN, GAL; YAAKOBI, ELAD; GRECHIKHIN, ELENA; DUENYAS, SHAHAF
To: SAMSUNG ELECTRONICS CO., LTD.
Reel/Frame 066752/0616 →
Continuity (1)
Related Publication 20250294261A1 · Sep 18, 2025
References Cited (26)
US 9792668B2 · Korkin · 2017 [cited by applicant]
US 10739189B2 · Pacala et al. · 2020 [cited by applicant]
US 10785422B2 · Ortiz Egea et al. · 2020 [cited by applicant]
US 11368603B2 · Kato · 2022 [cited by applicant]
US 11443550B2 · Ren et al. · 2022 [cited by applicant]
US 20230332954A1 · Lee et al. · 2023 [cited by applicant]
CN 107239781A · 2017 [cited by applicant]
KR 20160034897 · 2016 [cited by examiner]
WO 2017066825A1 · 2017 [cited by applicant]
WO 2018047171A1 · 2018 [cited by applicant]
WO 2022123047A1 · 2022 [cited by applicant]
Boaz Arad et al., “Hyperspectral Recovery from RGB”, (2016) V81, EMVA-YPA, The Interdisciplinary Computational Vision Laboratory, Ben-Gurion University of the Negev, 19 pages. [cited by applicant]
Naveed Akhtar et al., “Hyperspectral recovery from RGB images using Gaussian Processes”, Journal of LaTeX Class Files, vol. 14, No. 8, Aug. 2015, 15 pages, arXiv:1801.04654v2 [cs.CV] (Jul. 20, 2018). [cited by applicant]
Dilip K. Prasad et al., “Quick Approximation of Camera's Spectral Response from Casual Lighting”, 2013 IEEE International Conference on Computer Vision Workshops, pp. 844-851, DOI: 10.1109/ICCVW.2013.116. [cited by applicant]
Yi-Tun Lin et al., “Physically Plausible Spectral Reconstruction from RGB Images”, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pp. 2257-2266, DOI: 10.1109/CVPRW50498.2020.00274. [cited by applicant]
Radovan Slavuj et al., “To develop a method of estimating spectral reflectance from camera RGB values”, 2013 Colour and Visual Computing Symposium (CVCS), 6 pages, DOI: 10.1109/CVCS.2013.6626288. [cited by applicant]
Zhiyu Zhu et al., “Deep Amended Gradient Descent for Efficient Spectral Reconstruction from Single RGB Images”, 12 pages, arXiv:2108.05547v1 [eess.IV] (Aug. 12, 2021). [cited by applicant]
Berk Kaya et al., “Towards Spectral Estimation from a Single RGB Image in the Wild”, 10 pages, arXiv:1812.00805v1 [cs.CV] (Dec. 3, 2018). [cited by applicant]
Pai Chet Ng et al., “Efficient Hyperspectral Reconstruction from RGB Images with Line-Pixel Deconvolution”, 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop (IVMSP), 5 pages, DOI: 10.1109/IVM… [cited by applicant]
Nishchal Jagadeesha et al., “Skin Tone Assessment Using Hyperspectral Reconstruction from RGB Image”, 2023 15th International Conference on COMmunication Systems & NETworkS (COMSNETS), pp. 125-128, DOI: 10.1109/COMSNETS… [cited by applicant]
Emmanuel Martinez et al., “Efficient Transfer Learning for Spectral Image Reconstruction from RGB Images”, 2020 IEEE Colombian Conference on Applications of Computational Intelligence (IEEE ColCACI 2020), 6 pages, DOI: … [cited by applicant]
Yuqi Li et al., “Locally Linear Embedded Sparse Coding for Spectral Reconstruction From RGB Images”, IEEE Signal Processing Letters, vol. 25, No. 3, Mar. 2018 (published Nov. 21, 2017), pp. 363-367, DOI: 10.1109/LSP.201… [cited by applicant]
Hui Ding et al., “Smartphone based multispectral imager and its potential for point-of-care testing”, Analyst, 2019, vol. 144, No. 14, pp. 4380-4385, DOI: 10.1039/c9an00853e. [cited by applicant]
Annemarie McCarthy et al., “Low-Cost Multispectral Imager”, American Chemical Society, 2020, vol. 97, No. 10, pp. 3892-3898, DOI: 10.1021/acs.jchemed.0c00407. [cited by applicant]
Sewoong Kim et al., “Smartphone-based multispectral imaging and machine-learning based analysis for discrimination between seborrheic dermatitis and psoriasis on the scalp”, Biomedical Optics Express, 2019, vol. 10, No.… [cited by applicant]
Manu Parmar et al., “Selection of Optimal Spectral Sensitivity Functions for Color Filter Arrays”, IEEE Transactions on Image Processing, vol. 19, No. 12, 2010, Presented by Ji-Hoon Yoo from Color & Imaging Lab., Kyungp… [cited by applicant]