WO2023275583A1

WO2023275583A1 - Automated detection of gas emissions

Info

Publication number: WO2023275583A1
Application number: PCT/IB2021/000866
Authority: WO
Inventors: Elyes OUERGHI; Thibaud Ehret; Carlo DE FRANCHIS; Gabriele Facciolo; Thomas LAUVAUX; Jean-Michel Morel; Clément GIRON
Original assignee: Kayrros
Priority date: 2021-07-02
Filing date: 2021-12-10
Publication date: 2023-01-05

Abstract

A method of detecting a plume of a given gas composition from a hyperspectral image comprises obtaining a short-wave infrared (SWIR) image depicting an area of interest on a given date from an overhead image acquisition device, determining, for a plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas, applying an anomaly detection algorithm to the SWIR image to detect an anomaly therein that is attributable to a presence of the given gas within the area of interest, and identifying pixels as indicative of an emission of the given gas for those pixels of the SWIR image that are determined to be indicative of absorption of the given gas and determined to be indicative of the anomaly attributable to the presence of the given gas.

Description

AUTOMATED DETECTION OF GAS EMISSIONS

TECHNICAL FIELD

[0001] The present disclosure, in various embodiments, relates generally to a method for detecting gas plumes from hyper-spectral images. More particularly, the method relates to automatically detecting gas plumes from satellite short-wavelength infrared (SWIR) images.

BACKGROUND

[0002] The detection of large methane (CH4) leaks linked to oil and gas production plays a major role in the reduction of greenhouse gas (GHG) emissions. Over a period of 20 years, a methane molecule has a global warming potential 80 times larger than a carbon dioxide (CO2) molecule. A large portion of methane emissions could be controlled or avoided, as such emissions originate from oil rigs and other oil and gas infrastructures. In order to detect GHG fossil fuel emissions produced by human activities, several satellites have been placed in orbit around the Earth over the past ten years.

BRIEF SUMMARY

[0003] Aspects of the present disclosure relate to a method of detecting a plume of a given gas composition from a hyperspectral image. The method may comprise: obtaining a short-wave infrared (SWIR) image depicting an area of interest on a given date from an overhead image acquisition device, determining, for a plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas, applying an anomaly detection algorithm to the SWIR image to detect an anomaly therein that is attributable to a presence of the given gas within the area of interest, and identifying pixels as indicative of an emission of the given gas for those pixels of the SWIR image that are determined to be indicative of absorption of the given gas and determined to be indicative of the anomaly attributable to the presence of the given gas.

[0004] Certain preferred but non-limiting features of the method described above are the following, taken individually or in combination: [0005] the plurality of wavelengths for which the given gas exhibits absorption maximum comprises determining a plurality of absorption maximum of the given gas within a spectral band for which the given gas has a greatest absorption coefficient;

[0006] determining whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas comprises detecting absorption maxima in a negative logarithmic spectra of the at least one pixel that corresponds to absorption maxima in the absorption spectrum of the given gas;

[0007] determining, for the plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas comprises counting, for each pixel of the SWIR image, a number of absorption maximum exhibited;

[0008] the method further comprises, prior to determining whether the at least one pixel of the SWIR image is indicative of absorption of the given gas, removing from the SWIR image absorption of infrared light attributable to at least one of sun irradiance, albedo and atmosphere;

[0009] the method further comprises obtaining a plurality of SWIR images depicting the area of interest on a plurality of dates prior to the given date from the overhead image acquisition device, determining an average amount of absorption within the plurality of SWIR images attributable to the given gas, and removing from the SWIR image on the given date the average amount of absorption;

[00010] applying an anomaly detection algorithm to the SWIR image to detect an anomaly therein that is attributable to a presence of the given gas within the area of interest comprises: detecting an anomaly within the SWIR image by applying an anomaly detection algorithm to the SWIR image with a wavelength for which the given gas does not absorb (e.g., transmits) the given gas, detecting an anomaly within the SWIR image by applying an anomaly detection algorithm to the SWIR image with a wavelength for which the given gas absorbs the given gas, and identifying, within the SWIR image, pixels as indicative of the presence of the given gas within the area of interest for those pixels in which anomalies are detected in the SWIR image within the wavelength for which the given gas absorbs infrared light but are not detected as anomalies within the SWIR image within the wavelength for which the given gas does not absorb infrared light;

[00011] the anomaly detection algorithm may be a Reed-Xiaoli algorithm; and/or

[00012] the given gas being methane.

[00013] Another aspect of the present disclosure relates to a system configured to perform the method for detecting a plume of a given gas composition from a hyperspectral image as described herein.

[00014] The system comprises a processor communicatively coupled to at least one database from which a short-wave infrared image is obtained. The processor is configured to determine, for a plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas, apply an anomaly detection algorithm to the SWIR image to detect an anomaly therein that is attributable to a presence of the given gas within the area of interest, and identify pixels as indicative of an emission of the given gas for those pixels of the SWIR image that are determined to be indicative of absorption of the given gas and determined to be indicative of the anomaly attributable to the presence of the given gas.

[00015] Certain preferred but non-limiting features of the system described above are the following, taken individually or in combination:

[00016] the processor is further configured to detect an anomaly within the SWIR image by applying an anomaly detection algorithm to the SWIR image with a wavelength for which the given gas does not absorb the given gas, detect an anomaly within the SWIR image by applying an anomaly detection algorithm to the SWIR image with a wavelength for which the given gas absorbs the given gas, and identify, within the SWIR image, pixels as indicative of the presence of the given gas within the area of interest for those pixels in which anomalies are detected in the SWIR image within the wavelength for which the given gas absorbs infrared light but are not detected as anomalies within the SWIR image within the wavelength for which the given gas does not absorb infrared light;

[00017] the anomaly detection algorithm is a Reed-Xiaoli algorithm;

[00018] the plurality of wavelengths for which the given gas exhibits absorption maxima comprises determining a plurality of absorption maximum of the given gas within a spectral band for which the given gas has a greatest absorption coefficient; [00019] the processor is configured to determine whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas comprises the processor is configured to detect absorption maxima in a negative logarithmic spectra of the at least one pixel that corresponds to absorption maxima in the absorption spectrum of the given gas;

[00020] the processor is configured to determine, for the plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas comprises the processor is configured to count, for each pixel of the SWIR image, a number of absorption maximum exhibited;

[00021] the processor is further configured to remove from the SWIT image absorption of infrared light attributable to at least one of sun irradian, albedo, and atmosphere prior to the determination of whether the at least one pixel of the SWIR image is indicative of absorption of the given gas;

[00022] the processor is further configured to determine an average amount of absorption within a plurality of SWIR images that depict the area of interest on a plurality of dates prior to the given date obtained from the at least one database attributable to the given gas and configured to remove from the SWIR image on the given date the average amount of absorption; and/or

[00023] further comprising an interface at which an image of the emission of the given gas is output.

BRIEF DESCRIPTION OF DRAWINGS

[00024] FIG. 1 A illustrates an image depicting the concentration of a given gas within a plurality of pixels obtained from an overhead image acquisition device;

[00025] FIG. 1B illustrates an image depicting a number of spectrum maxima of the given gas counted for the pixels from the FIG. 1 A, which image was processed according to the method disclosed herein;

[00026] FIG. 2 depicts a flowchart of the method of the present disclosure;

[00027] FIG. 3 depicts plots an absorption coefficient for the given gas at different pressure and temperature conditions;

[00028] FIGS. 4A-4E depict an image at various stages of a process for detecting anomalies according to the method of the present disclosure; and

[00029] FIG. 5 is a schematic diagram of the general architecture of a system for performing the method of the present disclosure. DETAILED DESCRIPTION

[00030] The illustrations presented herein are not actual views of any particular component, device, or system, but are merely idealized representations employed to describe example embodiments of the present disclosure. The following description provides specific details of embodiments of the present disclosure in order to provide a thorough description thereof However, a person of ordinary skill in the art will understand that the embodiments of the disclosure may be practiced without employing many such specific details. Indeed, the embodiments of the disclosure may be practiced in conjunction with conventional techniques employed in the industry. In addition, only those process acts and systems necessary to understand the embodiments of the disclosure are described in detail below. Additional conventional acts and systems may be used. Also note, any drawings accompanying the application are for illustrative purposes only, and are thus not drawn to scale. Additionally, elements common between figures may have corresponding numerical designations.

[00031] As used herein, the terms “comprising,” “including,” “containing,” “characterized by,” and grammatical equivalents thereof are inclusive or open-ended terms that do not exclude additional, un-recited elements or method steps, but also include the more restrictive terms “consisting of,” “consisting essentially of,” and grammatical equivalents thereof.

[00032] As used herein, the term “may” with respect to a system, structure, feature, or method act indicates that such is contemplated for use in implementation of an embodiment of the disclosure, and such term is used in preference to the more restrictive term “is” so as to avoid any implication that other compatible systems, structures, features, and methods usable in combination therewith should or must be excluded.

[00033] As used herein, the singular forms following “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise.

[00034] As used herein, the term “and/or” includes any and all combinations of one or more of the associated listed items.

[00035] The present disclosure relates to a method of detecting plumes of a given gas, which method is illustrated in flowchart 200 of FIG. 2.

[00036] This method may comprise, at step 202, obtaining a short-wave infrared (SWIR) image depicting an area of interest on a given date from an overhead image acquisition device 502. [00037] Obtaining a SWIR image includes obtaining geo-located and radiometrically corrected top-of-atmosphere radiance data of SWIR bands on a given date from an overhead image acquisition device.

[00038] FIG. 1A illustrates an exemplary SWIR image 100 obtained at step 202. The SWIR image 100 shows a concentration of a gas emission of a commission of a given gas composition (here, methane) in parts per billion (shown by the scale on the image) within an area of interest.

[00039] Such SWIR image(s) may be obtained from a database 504.

[00040] The overhead image acquisition device may be a satellite. In some embodiments, the overhead image acquisition device may be the Sentinel-5P satellite or other satellite on which an infrared sensor is provided. The method may employ any infrared sensor that detects infrared light within a wavelength for which a given gas absorbs infrared light.

[00041] With particular reference to the Sentinel-5P satellite, the data (e.g., images) obtained maybe level 1 (L1) data. The Sentinel-5P satellite provides a dense spectrum comprising nearly 4,000 wavelengths for each pixel. This satellite obtains images for the entire Earth once a day. Thus, the method of the present disclosure may be used to detect gas emissions on a daily basis.

[00042] The SWIR portion of the electromagnetic spectrum comprises eight wavelength bands. Each band is composed of a given number of channels. A channel corresponds to an image at a particular wavelength. Thus, each band can be considered as a hyper-spectral image. In the present disclosure, the method may employ at least two active bands of infrared light for which the given gas absorbs infrared light. As used herein, the term “active band” refers to an infrared band having wavelengths (e.g., channels) of infrared light that are absorbed by the gas emission of a composition to be detected by the present method. When the given gas to be detected is composed of methane, the method may utilize SWIR bands 7 and 8, which encompass wavelengths in the range from about 2,300 nm to about 2,389 nm, inclusive. Within this range are primary absorption features of methane.

[00043] The method may, optionally, comprise obtaining geophysical quantities data in step 102. The geophysical quantities data may be derived from processing measurement data provided in the Level 1 data. Such geophysical quantities may be level 2 (L2) data from the Sentinel-5P satellite. Such geophysical quantities data may be obtained from a database 506. [00044] Such geophysical quantities data may comprise cloud maps, albedo, and X_CH4 column mixing ratios. Cloud maps may be utilized to assess useable pixels within the obtained images as clouds preclude gas emission detection. The XCH4 column mixing images may be used to identify gas emission plumes to validate the gas emission plume detected according to the present method.

[00045] The foregoing geophysical quantities data may provide a quantification of the gas emission in the atmosphere; however, such geophysical quantities data do not provide a plume detection.

[00046] The method may, optionally, comprise at step 104 pre-processing of the SWIR images obtained in step 102. Such pre-processing may include removing extraneous portions of the image, focusing the image on an area having a high gas concentration, and/or georeferencing the image.

[00047] The method may comprise, at step 202, obtaining an absorption spectrum from a database 508 and identifying a plurality of absorption maxima within at least a portion of the absorption spectrum for the gas emission. The database 508 may be, for example, the HITRAN spectral database, which comprises a compilation of spectroscopic parameters used to analyze the transmission and emission of light in various gaseous media. With reference to methane gas, the methane absorption spectrum varies as a function of pressure and temperature.

[00048] FIG. 3 illustrates absorption spectra for methane at multiple temperatures and pressures conditions. The absorption spectra in the plot of FIG. 3 plots absorption coefficients as a function of wavelength. Absorption coefficient maxima are identified for near-surface atmospheric conditions of 1 atm and 15°C (288K). A plurality of absorption maximum may be identified within a range of SWIR wavelengths for which SWIR images in step 102 may be obtained. For example, FIG. 3 illustrates with dots about 70 absorption maxima over wavelengths in a range from about 2300 nm to about 2389 nm, inclusive.

[00049] Accordingly, the method may comprise identifying a plurality of absorption maxima within a spectral band for which absorption of infrared light by a given gas is greatest.

[00050] The method may further comprise, in step 206, removing a background within the images obtained in step 202. FIG. 1B illustrates the SWIR image of FIG. 1A have the background has been removed.

[00051] In step 206, background removal may include removing contributions of albedo and atmosphere on the absorption of infrared light attributable to the gas emission. [00052] A simplified atmospheric absorption model is utilized to determine a value of each pixel of the image. For a given pixel P in the obtained image, the pixel is a vector in

where d is the number of channels in the obtained image. Each pixel component corresponds to a wavelength λ.

[00053] The simplified atmospheric absorption model takes into account the effect of sun irradiance F_I(λ), the albedo A, the absorption coefficients of the dry atmosphere K_atm(λ), water vapor

and methane In the formula (1) below,

denotes a thickness

of gas crossed by the infrared radiation before reaching the infrared sensor in the overhead image acquisition device, e_atm denotes a thickness of atmosphere crossed by the infrared radiation before reaching the infrared sensor in the overhead image acquisition device, and denotes a

thickness of water vapor crossed by the infrared radiation before reaching the infrared sensor in the overhead image acquisition device. Making implicit the dependence on wavelength l, the absorption model for the whole vector P is provided in formula (1).

[00054] In the absorption model, it is assumed that the albedo is roughly constant over the portion of the infrared spectrum utilized in the disclosed method as the albedo is extremely regular near 2000 nm. The absorption model also takes into consideration the absorption by the dry atmosphere, as a single gas whose absorption spectrum is well known. This spectrum includes absorption from methane that is always present in the atmosphere. represents the excess

of emitted methane over the one already present in the dry atmosphere.

[00055] The method utilizes — log(P) instead of P to create a linear model in which excesses of the given gas (e.g., methane gas emission) are positive values.

[00056] Background subtraction of step 206 may comprise a plurality of steps. Background subtraction removes the contribution of albedo and atmosphere from the spectrum for a given pixel. Background subtraction also sets the mean methane concentration to zero. As there is a nearly-constant concentration of methane within the atmosphere and large gas emissions, which are to be detected by the present method), rarely exceed 3% of this constant concentration.

[00057] In a first step of the background removal, albedo value that be known from geophysical quantities data obtained in step 102 may be used to remove the albedo component from each pixel. For a given pixel P₀ from a pre-processed hyperspectral image / obtained in step 102 and the albedo A corresponding to this given pixel, the albedo-corrected pixel is obtained by formula (2): P₁ = P₀ + log(A) (2)

[00058] The albedo may be assumed to be identical for each channel, as variations of albedo are minor in the infrared spectrum. After this first subtraction, P₁ contains only contributions from sun irradiance, atmosphere, water vapor, and CH₄.

[00059] For removing the contribution of the atmosphere, the method assumes that the irradiance and the absorption spectrum of the atmosphere are roughly constant over a short period of time (such as two weeks or less). Accordingly, the values of irradiance and atmosphere may be estimated from a time series of data. For each pixel P₁ observations X₁, ...X_n of the area of interest at earlier dates and without clouds. The background is then modeled as the principal component of observations X₁, ...X_n , which we denote F. To remove the background of P₁, the method comprises removing its projection on the subspace directed by F as shown in formula (3):

[00060] The foregoing removal of the contribution of the atmosphere may be optional as determining the contribution of the atmosphere relies upon a sufficiently sized time series. For example, the subtraction of the contribution of the atmosphere may only be performed for a given pixel P₁ if at least ten observations are obtainable for the time series. If a minimum number of observations cannot be obtained, the given pixel P₁ may be discarded. Further, if pixels from dates in the time series contain excesses of methane, the excess in these pixels may affect the principal component F and the background subtraction may remove a potential excess of methane in P₁. With a time series have a sufficient amount of data points, an excess of methane on one or two dates should not impact the principal component.

[00061] Furthermore, the observations X₁, ...X_n are obtained only from images in which less than 50% of pixels are affected by clouds.

[00062] The background subtraction step 106 may also comprise a step of equalizing (e.g., normalizing) a level of the given gas within the obtained SWIR image such that background of the given gas may be removed. Put differently, the method of the present disclosure is intended to detect gas emissions of a concentration that is greater than a concentration of the given gas that may exist naturally in the atmosphere or are due to gas emissions whether naturally occurring or synthetic that exists prior to the gas emission on the given date. Gas emissions detected may also be gas emissions that are man-made or unnatural and includes more intentional and unintentional gas emissions. [00063] The background subtraction works both spatially and spectrally.

[00064] With particular reference to methane, after the removal of albedo and atmosphere, only methane and water vapor should remain in the image.

[00065] To remove the background methane, the method comprises computing a spatial average (e.g., mean) of methane concentration M by projecting each pixel on the methane direction as shown in formula (4):

[00066] Subsequently, the computed mean M is subtracted from each pixel as shown in formula (5):

[00067] After this last operation, each pixel of the image displays a mix of water vapor and excess methane. Thus, detection of methane is possible when the concentration of water vapor is sufficiently low.

[00068] When the background is subtracted, even if the gas emission is only slightly visible in the image, the contrast between the gas emission and the background is significantly greater than prior to the background subtraction.

[00069] The method may further comprise, in step 208, counting a number of local maxima of the absorption spectra for each pixel of the obtained image from which background has been removed. The maxima are identified from the negative logarithmic spectrum of the pixels. Only those pixels having a maxima (e.g., a positive value in the negative logarithmic image) coinciding with a maxima in the methane absorption spectrum identified are counted in step 208.

[00070] At least one threshold may be applied in step 208 so as to reduce false positive detections of absorption maxima. A first threshold τ₁ (P) is the median of the spectrum of the pixel P. The first threshold prevents low absorption maxima from being detected. As only the highest absorption maxima of the given gas are selected in step 206, high maxima should be present in P.

[00071] A second threshold may be applied in step 208 that is adapted to the highest absorption maxima selected in step 206. For a wavelength L, the method comprises establishing a threshold τ₂ (λ) as the 70% quantile of all the values of the image at that wavelength. Put differently, for an image I, the 70% quantile of {P_¾ I P ∈ I} is utilized. In sum, for a maximum at wavelength A in a pixel P , the second threshold is max ( τ₁ (P), τ₂ (λ) ) .

[00072] A third threshold may be applied to distinguish excess gas emissions from background gas emissions based on the absorption maxima counted. The method may utilize an a contrario model. In this model, the a contrario assumption is that the SWIR image contains no excess CH₄, and compute the probability of false detection under this assumption. The method comprises indexing the selected absorption maxima i, going from 1 to 70 (in an embodiment in which 70 absorption maxima are selected), and denoting by ρ_i an empirical probability that an absorption maximum occurs at i in a “normal” image. If gas emission anomalies are presented in the obtained SWIR image, such anomalies are generally concentrated within a minority of pixels of the obtained image. Therefore, the method may comprise estimating ρ_i from the obtained SWIR image. This may result in a minor overestimation of the probability if some pixels have excess gas emissions. The random variable X_i, which is equal to 1 if the i-th maximum appears and 0 otherwise, follows a Bernoulli distribution with parameter ρ_i . To complete the a contrario model, it may be assumed that X₁, ...X_n are independent (in the absence of the gas emission. The method denotes by S (P) = X₁ +. . . +X_n the number of counted maxima on a given pixel P. Subsequently, a detection threshold t may be determined which guarantees a given false alarm rate p_ƒa

[00073] With reference to the selected absorption maxima, some absorption maxima may be given a different weight relative to other absorption maxima for the obtained image. For instance, with reference to methane gas detections, some maxima of the methane spectrum may not be observable due to the presence of other gases or particles in the atmosphere. By choosing the maxima for which the obtained image has more energy, it is possible to retain only the maxima are more likely to be observable.

[00074] The value of p_ƒa may be set to 10^-6, which amounts statistically to less than 0.01 false alarm per image.

[00075] The method may further comprise, in step 210, outputting an image for which the pixels are indicative of a number of absorption maxima as shown in FIG. 4A. In FIG. 4A, the plume of the given case to be detected by the method is encircled. White portions of FIGS. 4A- 4E are pixels excluded from the analysis due to, for example, cloud cover. [00076] The method may further comprise applying an algorithm for detecting anomalies on the obtained image. In steps 212 and 214, the image on which the anomaly detection algorithm is applied may be an image from which background has been removed as discussed in step 206.

[00077] In some embodiments, the Reed-Xiaoli algorithm is utilized to highlight those pixels indicative of an excess of the given gas and to remove anomalies that are not indicative of (e.g., attributable to) an excess of the given gas. The algorithm leams a model around a pixel of interest, then checks if this pixel follows the model.

[00078] For each pixel, the Reed-Xiaoli algorithm estimates a model from neighboring pixels. The algorithm assumes that all pixels in a neighborhood of the pixel under test (PUT) are independent and stem from the same random variable which follows a multivariate normal distribution. These pixels are used to compute the parameters of the normal law. Here, the neighborhood is a rectangular block centered on the PUT, deprived of a guard window. The size of the block is a parameter may be selectively chosen based on image resolution and on a size of the anomaly to be detected. Within the block centered on the PUT, pixels may be selected to be used to compute distribution parameters. The guard window may be s a smaller block centered on the PUT. Thus, Gaussian parameters may be computed with pixels outside the guard window. The guard window may be larger than the expected anomaly size.

[00079] The method may comprise determining an empirical mean m and covariance matrix C of the multivariate distribution with their usual non biased empirical estimators. Subsequently, the method comprises determining a likelihood of each pixel with respect to the multivariate distribution. Then, the likelihood of each pixel is compared to a detection threshold h which depends on m and C. The detection criterion on the likelihood for a pixel P can be written as follows:

Pixel P is an anomaly if (P — μ)^TC^_1(P — μ) > η.

[00080] Further, a detection threshold t may be selected which ensures a given false alarm rate p_ƒa. For a given threshold η, H₀ denotes the hypothesis that the pixel P follows the model and H₁ denotes the hypothesis that the pixel P is an anomaly (i.e., (P — μ)^TC^_1(P — μ) > η ) . The probability of false alarm is given by

By imposing a probability of false alarm per pixel, a corresponding detection threshold may be determined.

[00081] In step 212, the anomaly detection algorithm may be applied to detect anomalies for wavelengths for which the gas emission absorbs effectively no infrared light so as to obtain a first binary mask A₀ (FIG. 4B). The first binary mask A₀ is illustrated in grey pixels in FIG. 4B. [00082] In step 214, the anomaly detection algorithm may also be applied to detect anomalies for wavelengths for which the gas emission absorbs infrared light so as to obtain a second binary mask A_methane (FIG. 4C). FIG. 4C illustrates the second binary mask A_methane by grey pixels.

[00083] Anomalies in the second binary mask A_gas are most likely attributable to gas emission and anomalies in first binary mask A₀ are not attributable to the gas emission. There exists an overlap in anomalies identified by the first and second binary masks. Such anomalies are presumed to not be attributable to a gas emission.

[00084] In step 216, the method further comprises removing anomalies in the second binary mask A_methane that are present in the first binary mask A₀ to obtain a third binary mask (A_methane - A₀)⁺ (FIG. 4D) that comprises anomalies that should only be attributable to the gas emission. FIG. 4D illustrates the third binary mask (A_methane — A₀)⁺ in grey pixels.

[00085] Anomalies in A_methane are most likely due to methane and anomalies in A₀ are most likely not. However, some anomalies detected in A_methane are also detected in A₀. Therefore, it can be assumed that those anomalies are not due to methane. Indeed, methane has no effect on wavelengths used to compute A₀. Therefore, in the second binary mask A_methane, the anomalies detected in the first binary mask A₀ are removed.

[00086] In step 218, the method comprises combining the results of steps 210 and 216. More particularly, the method comprises identifying pixels as indicative of an emission of the given gas for those pixels of the at least one SWIR image that are determined to be indicative of absorption of the given gas in step 210 and determined to be indicative of anomalies attributable to the presence of the given gas within the area of interest in step 216.

[00087] In step 220, the method comprises outputting an image (FIG. 4E) depicting the plume of the given gas identified using the absorption maxima and the anomaly detection algorithm. In FIG. 4E, the detected plume is encircled.

[00088] While the method has been described with reference to processing of a single image and quantification of the emitted gas in the image, the method is not so limited. The method may be applied to a plurality of images. In some embodiments, the method may be applied to a time series of images such that plumes of the given gas may be detected over a period of time.

[00089] Embodiments of the present disclosure may comprise or utilize a special purpose or general-purpose computer including computer hardware, such as, for example, one or more processors and system memory, as discussed in greater detail below. Embodiments within the scope of the present disclosure also include physical and other computer-readable media for carrying or storing computer-executable instructions and/or data structures. In particular, one or more of the processes described herein may be implemented at least in part as instructions embodied in a non-transitory computer-readable medium and executable by one or more computing devices (e.g., any of the media content access devices described herein). In general, a processor (e.g., a microprocessor) receives instructions, from a non-transitory computer-readable medium, (e.g., a memory, etc.), and executes those instructions, thereby performing one or more processes, including one or more of the processes described herein.

[00090] Computer-readable media can be any available media that can be accessed by a general purpose or special purpose computer system. Computer-readable media that store computer-executable instructions are non-transitory computer-readable storage media (devices). Computer-readable media that carry computer-executable instructions are transmission media. Thus, by way of example, and not limitation, embodiments of the disclosure can comprise at least two distinctly different kinds of computer-readable media: non-transitory computer-readable storage media (devices) and transmission media.

[00091] Non-transitory computer-readable storage media (devices) includes RAM, ROM, EEPROM, CD-ROM, solid state drives (“SSDs”) (e.g., based on RAM), Flash memory, phase-change memory (“PCM”), other types of memory, other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store desired program code means in the form of computer-executable instructions or data structures and which can be accessed by a general purpose or special purpose computer.

[00092] Further, upon reaching various computer system components, program code means in the form of computer-executable instructions or data structures can be transferred automatically from transmission media to non-transitory computer-readable storage media (devices) (or vice versa). For example, computer-executable instructions or data structures received over a network or data link can be buffered in RAM within a network interface module (e.g., a “MC”), and then eventually transferred to computer system RAM and/or to less volatile computer storage media (devices) at a computer system. Thus, it should be understood that non- transitory computer-readable storage media (devices) can be included in computer system components that also (or even primarily) utilize transmission media.

[00093] Computer-executable instructions comprise, for example, instructions and data which, when executed at a processor, cause a general purpose computer, special purpose computer, or special purpose processing device to perform a certain function or group of functions. In some embodiments, computer-executable instructions are executed on a general-purpose computer to turn the general-purpose computer into a special purpose computer implementing elements of the disclosure. The computer executable instructions may be, for example, binaries, intermediate format instructions such as assembly language, or even source code. Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the described features or acts described above. Rather, the described features and acts are disclosed as example forms of implementing the claims.

[00094] Those skilled in the art will appreciate that the disclosure may be practiced in network computing environments with many types of computer system configurations, including, personal computers, desktop computers, laptop computers, message processors, hand-held devices, multi-processor systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, mainframe computers, mobile telephones, PDAs, tablets, pagers, routers, switches, and the like. The disclosure may also be practiced in distributed system environments where local and remote computer systems, which are linked (either by hardwired data links, wireless data links, or by a combination of hardwired and wireless data links) through a network, both perform tasks. In a distributed system environment, program modules may be located in both local and remote memory storage devices.

[00095] Embodiments of the present disclosure can also be implemented in cloud computing environments. In this description, “cloud computing” is defined as a model for enabling on-demand network access to a shared pool of configurable computing resources. For example, cloud computing can be employed in the marketplace to offer ubiquitous and convenient on-demand access to the shared pool of configurable computing resources. The shared pool of configurable computing resources can be rapidly provisioned via virtualization and released with low management effort or service provider interaction, and then scaled accordingly.

[00096] A cloud-computing model can be composed of various characteristics such as, for example, on-demand self-service, broad network access, resource pooling, rapid elasticity, measured service, and so forth. A cloud-computing model can also expose various service models, such as, for example, Software as a Service (“SaaS”), Platform as a Service (“PaaS”), and Infrastructure as a Service (“IaaS”). A cloud-computing model can also be deployed using different deployment models such as private cloud, community cloud, public cloud, hybrid cloud, and so forth. In this description and in the claims, a “cloud-computing environment” is an environment in which cloud computing is employed.

[00097] FIG. 5 illustrates a block diagram of an example system 500 that may be configured to perform one or more of the processes described above. The system 500 is communicatively coupled to a plurality of databases. The plurality of databases may include at least one database 504 from which SWIR images are obtained, at least one database 506 from which geophysical quantities data is obtained, and at least one database 508 from which an absorption spectrum is obtained. The plurality of databases may obtain images and other data from overhead image acquisition devices. In some embodiments, the overhead image acquisition devices may be different satellites comprising different sensors and image acquisitions devices.

[00098] The databases 504, 506, and 508 may be stored on a non-transitory computer- readable storage media (device) as previously described herein. As shown by FIG. 5, the system 500 may comprise a computing device 526. The computing device 526 may comprise a processor 528, a memory 530, a storage device 532, an I/O interface 534, and a communication interface 536, which may be communicatively coupled by way of a communication infrastructure 538. While an example computing device 526 is shown in FIG. 5, the components illustrated in FIG. 5 are not intended to be limiting. Additional or alternative components may be used in other embodiments. Furthermore, in certain embodiments, the computing device 526 can include fewer components than those shown in FIG. 5. Components of the computing device 526 shown in FIG. 5 will now be described in additional detail.

[00099] In one or more embodiments, the processor 528 includes hardware for executing instructions, such as those making up a computer program. By way of non-limiting example, to execute instructions, the processor 528 may retrieve (or fetch) the instructions from an internal register, an internal cache, the memory 530, or the storage device 532 and decode and execute the instructions. In one or more embodiments, the computing device 526 may include one or more internal caches for data, instructions, or addresses. By way of non-limiting example, the computing device 526 may include one or more instruction caches, one or more data caches, and one or more translation look aside buffers (TLBs). Instructions in the instruction caches may be copies of instructions in the memory 530 or the storage 532.

[000100] The computing device 526 includes memory 530, which is coupled to the processor 528. The memory 530 may be used for storing data, metadata, and programs for execution by the processor(s) 528. The memory 530 may include one or more of volatile and non- volatile memories, such as Random Access Memory (“RAM”), Read Only Memory (“ROM”), a solid state disk (“SSD”), Flash, Phase Change Memory (“PCM”), or other types of data storage. The memory 530 may be internal or distributed memory.

[000101] The computing device 526 includes the storage device 532 that includes storage for storing data or instructions. By way of non-limiting example, storage device 532 can comprise a non-transitory storage medium described above. The storage device 532 may include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disc, a magneto-optical disc, magnetic tape, or a Universal Serial Bus (USB) drive or a combination of two or more of these. The storage device 532 may include removable or non-removable (or fixed) media, where appropriate. The storage device 532 may be internal or external to the computing device 526. In one or more embodiments, the storage device 532 is non-volatile, solid-state memory. In other embodiments, the storage device 532 includes read-only memory (ROM). Where appropriate, this ROM may be mask programmed ROM, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), electrically alterable ROM (EAROM), or flash memory or a combination of two or more of these.

[000102] The computing device 526 also includes one or more input or output (“I/O”) devices/interfaces 534, which are provided to allow a user to provide input to, receive output from, and otherwise transfer data to and receive data from computing device 526. The I/O devices/interfaces 534 may include a mouse, a keypad or a keyboard, a touch screen, a camera, an optical scanner, network interface, modem, other known I/O devices or a combination of such I/O device/interfaces. The touch screen may be activated with a stylus or a finger.

[000103] The I/O devices/interfaces 534 may include one or more devices for presenting output to a user, including, but not limited to, a graphics engine, a display (e.g., a display screen), one or more output drivers (e.g., display drivers), one or more audio speakers, and one or more audio drivers. In certain embodiments, the I/O interface 534 is configured to provide graphical data to a display for presentation to a user. The graphical data may be representative of one or more graphical user interfaces and/or any other graphical content as may serve a particular implementation.

[000104] The computing device 526 can further include a communication interface 328. The communication interface 536 can include hardware, software, or both. The communication interface 536 can provide one or more interfaces for communication (such as, for example, packet- based communication) between the computing device 526 and one or more other computing devices or networks. As an example and not by way of limitation, the communication interface 536 may include a network interface controller (NIC) or network adapter for communicating with an Ethernet or other wire-based network or a wireless MC (WNIC) or wireless adapter for communicating with a wireless network, such as a WI-FI. The computing device 526 can further include a bus. The bus can comprise hardware, software, or both that couples components of computing device 526 to each other.

[000105] While the present disclosure has been described herein with respect to certain illustrated embodiments, those of ordinary skill in the art will recognize and appreciate that it is not so limited. Rather, many additions, deletions, and modifications to the illustrated embodiments may be made without departing from the scope of the invention as hereinafter claimed, including legal equivalents thereof. In addition, features from one embodiment may be combined with features of another embodiment while still being encompassed within the scope of the invention as contemplated by the inventors.

Claims

CLAIMS What is claimed is:

1. A method of detecting a plume of a given gas composition from a hyperspectral image, comprising: obtaining a short-wave infrared (SWIR) image depicting an area of interest on a given date from an overhead image acquisition device; determining, for a plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas; applying an anomaly detection algorithm to the SWIR image to detect an anomaly therein that is attributable to a presence of the given gas within the area of interest; and identifying pixels as indicative of an emission of the given gas for those pixels of the SWIR image that are determined to be indicative of absorption of the given gas and determined to be indicative of the anomaly attributable to the presence of the given gas.

2. The method of claim 1, wherein applying an anomaly detection algorithm to the SWIR image to detect an anomaly therein that is attributable to a presence of the given gas within the area of interest comprises: detecting an anomaly within the SWIR image by applying an anomaly detection algorithm to the SWIR image with a wavelength for which the given gas does not absorb the given gas; detecting an anomaly within the SWIR image by applying an anomaly detection algorithm to the SWIR image with a wavelength for which the given gas absorbs the given gas; and identifying, within the SWIR image, pixels as indicative of the presence of the given gas within the area of interest for those pixels in which anomalies are detected in the SWIR image within the wavelength for which the given gas absorbs infrared light but are not detected as anomalies within the SWIR image within the wavelength for which the given gas does not absorb infrared light.

3. The method of either of claims 1 or 2, wherein the anomaly detection algorithm is a Reed-Xiaoli algorithm.

4. The method any of claims 1 -3, wherein the plurality of wavelengths for which the given gas exhibits absorption maxima comprises determining a plurality of absorption maximum of the given gas within a spectral band for which the given gas has a greatest absorption coefficient.

5. The method of any of claims 1-4, wherein determining whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas comprises detecting absorption maxima in a negative logarithmic spectra of the at least one pixel that corresponds to absorption maxima in the absorption spectrum of the given gas.

6. The method of any of claims 1-5, wherein determining, for the plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas comprises counting, for each pixel of the SWIR image, a number of absorption maximum exhibited.

7. The method of any of claims 1 -6, further comprising, prior to determining whether the at least one pixel of the SWIR image is indicative of absorption of the given gas, removing from the SWIR image absorption of infrared light attributable to at least one of sun irradiance, albedo and atmosphere.

8. The method of any of claims 1 -7, further comprising obtaining a plurality of SWIR images depicting the area of interest on a plurality of dates prior to the given date from the overhead image acquisition device, determining an average amount of absorption within the plurality of SWIR images attributable to the given gas, and removing from the SWIR image on the given date the average amount of absorption.

9. The method of any of claims 1-8, further comprising outputting an image of the emission of the given gas.

10. The method of any of claims 1-9, wherein the given gas being methane.

11. A system for detecting a plume of a given gas composition from a hyperspectral image, comprising: a processor communicatively coupled to at least one database from which a short-wave infrared image is obtained, the processor configured to: determine, for a plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas; apply an anomaly detection algorithm to the SWIR image to detect an anomaly therein that is attributable to a presence of the given gas within the area of interest; and identify pixels as indicative of an emission of the given gas for those pixels of the SWIR image that are determined to be indicative of absorption of the given gas and determined to be indicative of the anomaly attributable to the presence of the given gas.

12. The system of claim 11 , wherein the processor is further configured to: detect an anomaly within the SWIR image by applying an anomaly detection algorithm to the SWIR image with a wavelength for which the given gas does not absorb the given gas; detect an anomaly within the SWIR image by applying an anomaly detection algorithm to the SWIR image with a wavelength for which the given gas absorbs the given gas; and identify, within the SWIR image, pixels as indicative of the presence of the given gas within the area of interest for those pixels in which anomalies are detected in the SWIR image within the wavelength for which the given gas absorbs infrared light but are not detected as anomalies within the SWIR image within the wavelength for which the given gas does not absorb infrared light.

13. The system of either of claims 11 or 12, wherein the anomaly detection algorithm is a Reed-Xiaoli algorithm.

14. The system any of claims 11-13, wherein the plurality of wavelengths for which the given gas exhibits absorption maxima comprises determining a plurality of absorption maximum of the given gas within a spectral band for which the given gas has a greatest absorption coefficient.

15. The system of any of claims 11-14, wherein the processor is configured to determine whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas comprises the processor is configured to detect absorption maxima in a negative logarithmic spectra of the at least one pixel that corresponds to absorption maxima in the absorption spectrum of the given gas.

16. The system of any of claims 11-15, wherein the processor is configured to determine, for the plurality of wavelengths for which a given gas exhibits an absorption maximum, whether at least one pixel of the SWIR image is indicative of absorption of infrared light within the given gas comprises the processor is configured to count, for each pixel of the SWIR image, a number of absorption maximum exhibited.

17. The system of any of claims 11-16, wherein the processor is further configured to remove from the SWIT image absorption of infrared light attributable to at least one of sun irradian, albedo, and atmosphere prior to the determination of whether the at least one pixel of the SWIR image is indicative of absorption of the given gas.

18. The system of any of claims 11-17, wherein the processor is further configured to determine an average amount of absorption within a plurality of SWIR images that depict the area of interest on a plurality of dates prior to the given date obtained from the at least one database attributable to the given gas and configured to remove from the SWIR image on the given date the average amount of absorption.

19. The system of any of claims 11-18, further comprising an interface at which an image of the emission of the given gas is output.