Granted Patent B2

US 12,270,758 · App. 17/970,337 · Granted Apr 8, 2025

Inventors: Edouard Machover (Paris, FR); Gabriele Facciolo (Cachan, FR); Jean-Michel Morel (Paris, FR); Carlo De Franchis (Paris, FR); Thibaud Ehret (Champigny sur Marne, FR); Aurélien De Truchis (Paris, FR); Matthieu Mazzolini (Paris, FR); Thomas Lauvaux (Montpellier, FR)

Assignee: Kayrros

G01N21/3504G06T5/70G06T5/77G06V10/60G06V20/13G06T2207/10036G06T2207/10048

Abstract

Claims (47)

1. A method, comprising:

determining an effective image of an area of interest, based on an imaging acquisition performed by an imager installed on-board an Earth observation satellite;

estimating a reference image of the area of interest, wherein the reference image corresponds to the effective image in the absence of a gas emission located in the area of interest;

normalizing the effective image with the reference image;

determining a gas emission mask based on the normalized effective image; and

correlating an amount of light intensity within the gas emission mask to a gas concentration of a gas emission located in the area of interest.

2. The method of claim 1 , wherein the effective image is a single band image.

3. The method of claim 1 , wherein the imager is a multispectral imager configured to capture a plurality of single band images of the area of interest and wherein determining the effective image comprises calculating a ratio between a first and a second single band images captured by the multispectral imager.

4. The method of claim 3 , wherein the first single band image covers a first spectral band in which the gas emission at least partially absorbs light, and wherein the second single band image covers a second spectral band in which the gas emission is absorbed to a lesser extent than in the first spectral band.

5. The method of claim 3 , wherein determining the effective image further comprises applying a logarithm to the calculated ratio.

6. The method of claim 1 , wherein the imager is a multispectral imager configured to capture a plurality of single band images of the area of interest and wherein determining the effective image comprises calculating a linear combination of logarithms of each single band image captured by the imager.

7. The method of claim 6 , wherein coefficients of the logarithms of single band images more affected by the gas emission are positive and coefficients of the logarithms of the single band images less affected by the gas emission are negative.

8. The method of any of claim 7 , wherein the coefficients of the logarithms are computed by an optimization algorithm based on the correlation of bands and on the gas absorption spectra so as to maximize a signal-to-noise ratio of the effective image.

9. The method of claim 1 , wherein estimating the reference image of the area of interest comprises performing image reconstruction based on the effective image.

10. The method of claim 9 , wherein performing image reconstruction based on the effective image comprises applying a non-local means denoising algorithm to the effective image.

11. The method of claim 1 , wherein estimating the reference image of the area of interest comprises a tensor decomposition of a tensor composed of images of a time-series of images of the area of interest.

12. The method of claim 11 , wherein estimating the reference image of the area of interest comprises projecting the effective image on the time-series of images of the area of interest.

13. The method of claim 12 , wherein projecting the effective image on the time-series of images of the area of interest results in an intermediate reference image having pixels each associated to a projection residual and further comprising:

discarding pixels of the intermediate reference image based on their projection residuals;

removing pixels of the effective image and of the time-series of images in correspondence with the discarded pixels;

projecting the effective image with removed pixels on the time-series of images with removed pixels to determine weights of a linear combination of the images of the time-series;

calculating the reference image, based on the determined weights.

14. The method of claim 12 , wherein said projecting comprises performing one projection for each region of similar albedo in the effective image and the time-series of images.

15. The method of claim 1 , wherein determining the gas emission mask comprises thresholding the normalized effective image.

16. The method of claim 1 , wherein correlating an amount of light intensity within the gas emission mask to a gas concentration of a gas emission located in the area of interest comprises correlating the amount of light intensity in each pixel of the gas emission mask to a dimming function of the imager and determining an optical path length of the gas emission for each pixel.

17. The method of claim 16 , further comprising calculating a concentration of the gas emission from the determined optical path length.

18. A method, comprising:

determining a time-series of effective images of an area of interest, based on several imaging acquisitions performed by an imager installed on-board an Earth observation satellite;

for each effective image of the time-series:

estimating a reference image of the area of interest, wherein the reference image corresponds to the effective images in the absence of a gas emission located in the area of interest;

normalizing the effective image with the reference image;

determining a gas emission mask based on the normalized effective image; and

correlating an amount of light intensity within the gas emission mask to a gas concentration of a gas emission located in the area of interest;

monitoring methane pollution based on the gas concentration over the time-series of effective images.

19. A data processing apparatus, comprising a processor configured to implement:

determining an effective image of an area of interest, based on an imaging acquisition performed by an imager installed on-board an Earth observation satellite;

estimating a reference image of the area of interest, wherein the reference image corresponds to the effective image in the absence of a gas emission located in the area of interest;

normalizing the effective image with the reference image;

determining a gas emission mask based on the normalized effective image; and

correlating an amount of light intensity within the gas emission mask to a gas concentration of a gas emission located in the area of interest.

20. A non-transitory computer medium, comprising:

code instructions stored on the non-transitory computer medium, wherein when executed by a processor, the code instructions are configured to cause the processor to implement:

determining an effective image of an area of interest, based on an imaging acquisition performed by an imager installed on-board an Earth observation satellite;

estimating a reference image of the area of interest, wherein the reference image corresponds to the effective image in the absence of a gas emission located in the area of interest;

normalizing the effective image with the reference image;

determining a gas emission mask based on the normalized effective image; and

correlating an amount of light intensity within the gas emission mask to a gas concentration of a gas emission located in the area of interest.

Assignments (1)

Continuity (3)

References Cited (15)