WO2024008522A1

WO2024008522A1 - Generating a spectrum of a sample

Info

Publication number: WO2024008522A1
Application number: PCT/EP2023/067552
Authority: WO
Inventors: Michael Hanke
Original assignee: Trinamix Gmbh
Priority date: 2022-07-05
Filing date: 2023-06-28
Publication date: 2024-01-11

Abstract

A computer-implemented method for generating a spectrum of a sample comprising: - receiving a plurality of spectral data sets suitable for obtaining a plurality of spectra, - aggregating at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra, and - providing the at least one aggregated spectrum.

Description

Generating a spectrum of a sample

The present invention is in the field of generating a spectrum of a sample. In particular, it relates to a computer-implemented method for generating a spectrum of a sample, a spectrometer device, a computer program including instructions for executing steps of the method as described herein, a computer-readable data medium storing a computer program and a use of at least one aggregated spectrum obtained by the method according to any one of the preceding claims for determining material information.

Background

Spectroscopy allows the analysis of the chemical composition of a sample. The sample is allowed to interact with electromagnetic radiation. From the reflected electromagnetic radiation a spectrum is obtained. Especially for portable spectrometers which are used under conditions far away from controlled laboratory conditions, various factors can have an influence on the spectrum. For example, when moving the spectrometer device by hand, the distance to the sample varies, the device is moved along the sample and at some point, it exceeds the limit of the sample or some disturbing pieces are in the sample, for example a leaf in a grape harvest. This leads to undesired spectral information, which need to be removed in a lengthy manual analysis.

It was hence the object of the present invention to overcome these shortcomings. In particular, a method for a robust and reproducible generation of a spectrum with a high signal-to-noise-ratio, without undesired artefacts and without contamination from an undesired sample is seeked for.

Summary

These objects were achieved by the present invention. In one aspect it relates to a computer- implemented method for generating a spectrum of a sample comprising: receiving a plurality of spectral data sets suitable for obtaining a plurality of spectra, aggregating at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra, and providing the at least one aggregated spectrum.

In another aspect it relates to a spectrometer device comprising: a detector configured to generate a plurality of spectral data sets, a controller configured to provide the plurality of spectral data sets to a processor for carrying out the steps according to the method with all the embodiments as described herein. In another aspect it relates to a computer program including instructions for executing steps of the method according to any one of the preceding claims.

In another aspect it relates to non-transitory computer-readable data medium storing a computer program.

In another aspect it relates to a use of the at least one aggregated spectrum obtained by the method according to any one of the preceding claims for determining material information.

In another aspect it relates to a system for generating a spectrum of a sample comprising: an input for receiving a plurality of spectral data sets suitable for obtaining a plurality of spectra, a processor for aggregating at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra, and an output for providing the at least one aggregated spectrum.

The present invention provides means for an efficient, fast and robust generation of a spectrum of a sample. Especially in scenarios where a spectrometer is used far away from standardized laboratory condition a spectrum resulting from such a measurement may not provide the information desired and/or more than one part of a sample may be measured. By measuring these spectra and including them into the analysis, results with low quality may be obtained. Low-quality results and/or low-quality spectra may comprise spectra with undesired artefacts and/or undesired contamination and/or with a low signal-to-noise ratio. High-quality results and/or high-quality spectra may comprise spectra without undesired artefacts and/or undesired contamination and/or with a high signal-to-noise ratio. High-quality spectra can be obtained by generating a spectrum of a sample as described herein. This can be facilitated by aggregating a plurality of spectra. The user is released of the burden to analyze the data himself in order to decide for undesired spectral data. Followingly, generating a spectrum from a sample including an aggregation step saves time during the data evaluation. Furthermore, undesired spectral data sets can be identified and for example be eliminated and/or remeasured. Thereby, even more time of the user can be saved since no further measurement series needs to be initiated when having processed the data already on the go. The user can just continue measuring the undesired spectrum while still being at the point of interest. In addition, data storage may be used more efficiently by avoiding to safe undesired spectral data sets. Compressing data as facilitated by the invention described herein by means of aggregation, reduces the amount of data to be transferred leading to a faster and less resource intensive data transfer. Excluding the undesired spectral data sets from aggregation provides an increased signal-to-noise-ratio of the spectrum, thereby facilitate extracting chemical information in an improved manner. Aggregating spectra into more than one aggregated spectrum provides the user with more than one spectrum worth to analyze due to its high quality corresponding to more than one different chemical composition. By doing so, more than one result may be obtained from one measurement series. Overall, generating a spectrum from a sample as described herein provides more accurate and/or precise spectral results because of the artifact-free and contamination-free spectra with high signal-to-noise ratio. High signal-to-noise ratio may refer to a signal-to-noise ratio larger than a threshold value. Low signal-to-noise ratio may refer to a signal-to-noise ratio smaller than or equal to a threshold value. High signal-to-noise ratio may refer to a signal-to-noise ratio larger than a threshold value. Low signal-to-noise ratio may refer to a signal-to-noise ratio smaller than or equal to a threshold value. Ultimately, this opens the opportunity to apply spectroscopic methods to users with non-expert knowledge to operate a spectrometer since the measured data may be compensated for disturbances.

Any disclosure and embodiments described herein relate to the methods, the systems, the use, the computer program element, the non-transitory computer-readable data medium lined out above and vice versa. Advantageously, the benefits provided by any of the embodiments and examples equally apply to all other embodiments and examples and vice versa. The systems may be suitable for carrying out the steps of the methods.

Spectral data set comprises information associated with spectroscopy measurement result. Spectroscopy may be performed with a spectrometer. Spectroscopy may examine the interaction of light with matter as a function of a measure suitable for suitable for expressing a defined energy value of the light. Spectroscopical methods are known in the art [Prof. Dr. Gunter Gauglitz, Dr. David S. Moore; Handbook of Spectroscopy: Second, Enlarged Edition; 02.04.2014; WILEY-VCH Verlag GmbH & Co., Weinheim, Germany; Online ISBN:9783527654703], Examples for spectroscopical methods may be absorption spectroscopy, emission spectroscopy, analysis of elastic scattering and/or inelastic scattering, reflection spectroscopy, impedance spectroscopy, resonance spectroscopy, nuclear spectroscopy. In particular, IR spectroscopy, in particular MIR, NIR and/or FTIR spectroscopy, UWis spectroscopy, fluorescence spectroscopy and raman spectroscopy.

Spectral data set may comprise data suitable for deriving a spectrum and/or a spectrum and/or data derived from a spectrum. In particular spectral data set may be an IR spectral data set and/or may comprise data suitable for deriving an IR spectrum and/or an IR spectrum and/or data derived from an IR spectrum. Data suitable for deriving a spectrum may comprise an interferogram and/or any similar prestage of a spectrum. A spectrum may be determined based on a spectral data set, in particular wherein the spectral data set may comprise a prestage of a spectrum and/or does not comprise a spectrum. A spectrum may be derived from a spectral data set by applying a mathematical operation, e.g. a Fourier transform or similar transformations as known in the art. A spectrum may relate light intensity and/or any measure derived from the intensity to a measure for the energy of light during the measurement. Measures for the energy of light may be energy values, e.g. given in eV or J, and/or wavelengths, e.g. given in m, and/or wavenumbers, e.g. given in reciprocal m, and/or frequency, e.g. given in ^{s 1} and/or any other measure suitable for expressing a defined energy value of the light. Measures derived from light intensity may comprise but are not limited to relative intensities by relating the final intensity to the initial intensity, absorbance, extinction or the like. The exact measure used depends on the type of spectroscopy and the use case for which the method is applied. A part of a spectrum may be a spectrum. Such a spectrum may have a lower range for the measure for the energy of the light. Hence, when referring to a plurality of spectra a plurality of parts of spectra may be understood. A spectrum may be an optical spectrum. Spectrum may comprise normalized spectrum and/or spectra suitable for being normalized. Spectrum may comprise a derivative, antiderivative and/or transformation of a spectrum. Plurality of spectral data sets may be time dependent and thus, may comprise a time series. Followingly, at least two of the plurality of spectral data sets may be generated at different points in time. Spectral data sets may be generated at points in time separated by a constant time interval and/or varying time intervals. Spectral data sets may be generated with the same or different spectrometer devices. In some embodiments, at least a part of the plurality of spectral data sets may be generated with a spectrometer device as described herein and another part of the plurality of spectral data sets may be generated with spectrometer device different than the first one, eventually the second or any other spectrometer device may be constructed as described herein.

A difference among the plurality of spectra as used herein is a broad term and is to be given its ordinary and customary meaning to a person of ordinary skill in the art and is not to be limited to a special or customized meaning. Difference may refer to at least one data point different among the spectra. In particular, difference may referto a difference exceeding the precision of a measurement. Precision in a measurement depends on the hardware setup used for performing spectroscopical methods. Precision may be determined by reference measurements, e.g. with a more accurate analytical method and/or device. Precision of a spectroscopical method may be known by a skilled person performing spectroscopy and known in corresponding literature. Precision may refer to a deviation of a measurement result determined by spectroscopical methods from the mean value of results if measurement is conducted several times. Hence a difference may be determined if the result of at least one measurement and thus, spectral data set, deviates 5% or more from the mean value of results for this sample, in particular when specifically accurate measurement setups may be used when the result of at least one measurement and thus, spectral data set, deviates 3% or more from the mean value of results for this sample. Difference among the plurality of spectra may be determined by comparing the plurality spectral data sets, in particular the plurality of spectra. Difference among the plurality of spectra may be determined based on a feature. The feature may be a distinguishing. The difference among the plurality of spectra may be determined based a feature included in only a part of the plurality of spectra.

For example, the distinguishing feature may be an additional peak, a different intensity compared to at least one of the plurality of spectra, a different peak position compared to at least another one of the plurality of spectra, a different peak shape compared to at least another one of the plurality of spectra, a different multiplicity of a peak, e.g. singlet, doublet or the like, compared to at least another one of the plurality of spectra and/or a different peak baseline on the right and/or left side (e.g. downfield and/or upfield side) of the peak compared to at least another one of the plurality of spectra. Peak shape may refer to the peak width, in particular the full width at half maximum (FWHM), Lorentzian or Gaussian peak shape and/or slope of the peak. The feature may be suitable to determine a difference in material information of the sample.

It is to be understood that referring to any of the mentioned measure for the energy of light any measure mentioned before and/or any measure equally suitable for expressing and/or referring to an energy of light may be understood as placed additionally and/or alternatively in this document. Further, a “plurality” may refer to at least two, in particular at least 5, 50, 100 or 500. Furthermore, when referring to “light” herein, it is to be understood that the term light is not limited to visible light, but may include light with all wavelengths , in particular light in the range between 1 O’ ¹¹ m and 10’⁴ m such as infrared light, visible light and/or ultraviolet light. The word light may be used interchangeably with the term electromagnetic radiation.

The term “detector” as used herein is a broad term and is to be given its ordinary and customary meaning to a person of ordinary skill in the art and is not to be limited to a special or customized meaning. The term specifically may refer, without limitation, to a device which is configured for generating a plurality of spectral data set from a sample. The detector specifically may be or may comprise at least one photodetector configured for generating data related to material information of a sample. The detector specifically may comprise at least one photosensitive area. More specifically, the detector may comprise at least one detector array. As an example, one or more filters, such as single wavelength bandpass filters and/or an array of bandpass filters and/or a length variable filter, may be disposed on top of the detector. More than one photosensitive area may be comprised in the detector. Consequently, each photosensitive area of the detector may detect a different spectral portion of the incident electromagnetic radiation. Other embodiments, however, are also feasible. A “photosensitive area” generally refers to an area of the optical sensor which may be illuminated externally by electromagnetic radiation and may in response to illumination generate at least one sensor signal. The sensor signal may be suitable for determining an intensity of the light and/or a measure rel. The photosensitive area may specifically be located on a surface of the respective optical sensor. Other embodiments, however, are feasible. The optical sensor specifically may be or may comprise photodetectors, preferably inorganic photodetectors, more preferably inorganic semiconductor photodetectors, most preferably silicon photodetectors. Specifically, the optical sensors may be or may comprise inorganic photodetectors which are sensitive in the infrared spectral range, preferably in the range of 780 nm to 3.0 micrometers. Specifically, the optical sensors may be sensitive in the part of the near infrared region where silicon detectors are applicable specifically in the range of 700 nm to 1000 nm. Infrared optical sensors which may be used for optical sensors may be commercially available infrared optical sensors, such as infrared optical sensors commercially available under the brand name Hertz- stueck™ from trinamiX GmbH, D-67056 Ludwigshafen am Rhein, Germany. Thus, as an example, the optical sensors may comprise at least one optical sensor, more preferably at least one photodetector selected from the group consisting of: a Ge detector, a Si detector, a GaAs detector, an InGaAs detector, an extended InGaAs detector, an InAs detector, an InSb detector, a HgCdTe detector, a Ge:Au detector, a Ge:Hg detector, a Ge:Cu detector, a Ge:Zn detector, a Si:Ga detector, a Si:As detector, a PbS detector. Depending on the specific electromagnetic radiation to detect, the material of the detector may be chosen. A wide variety of detector for different use cases are known in the art. For example, PbS detector may be especially suitable for detecting infrared light, in particular near-infrared light. Si detector may be suitable for detecting light of wavelengths between 200 nm and 1000 nm, thereby detecting light in the visible and the infrared range. Some Si detectors may be suitable for detecting light in the range of 0.07 nm and 1100 nm, thereby opening the application towards the detection of ultraviolet and/or x-ray light.

Additionally or alternatively, the optical sensors may comprise at least one bolometer, preferably a bolometer selected from the group consisting of a VO bolometer and an amorphous Si bolometer. A plurality of photosensors may be arranged in a matrix. The matrix may be composed of independent optical sensors. Thus, a matrix may be composed of inorganic photodiodes. Detector may be or may comprise a photosensitive device. Photosensitive device may be configured to generate one output signal that may be suitable for determining an intensity of light. Photosensitive device suitable for generating two or more output signals, for example at least one CCD and/or CMOS device, may be referred to as two or more detectors. Additionally or alternatively, however, a detector may be a CCD detector, such as a CCD detector chip, and/or a CMOS detector, such as a CMOS detector chip. For example, a CCD detector may be suitable for detecting visible light (350 nm to 700 nm).

Controller is operatively coupled to the processor and/or detector. The controller may be coupled via a wired and/or wireless connection such as one of ethernet, USB, LAN, WLAN, Bluetooth and the like. The controller may comprise an interface. Such an interface may be suitable for receiving and/or providing spectra and/or data related to spectra. Interface may comprise one or more of serial or parallel interfaces or ports, USB, Centronics Port, FireWire, HDMI, Ethernet, Bluetooth, RFID, Wi-Fi, USART, or SPI, or analogue interfaces or ports such as one or more of ADCs or DACs, or standardized interfaces or ports to further devices. The connection may be further suitable for receiving an indication of the at least one difference. The controller may be suitable for receiving and/or providing a plurality of spectral data sets. Furthermore, the controller may be suitable for controlling the functioning of the detector and/or processor. The controller may be suitable for generating a control signal. The control signal may be suitable for operating the detector and/or processor. Operating the detector may include initiating measurement of a plurality of spectral data sets. Operating the processor may include initiating processing of a plurality of spectra, e.g. aggregating a plurality of spectra. In some embodiments, the term may refer, without limitation, to a device or combination of devices capable and/or configured for performing at least one computing operation and/or for controlling at least one function of at least one other device, such as of at least one other component of the portable spectrometer device. Specifically, the at least one controller may be embodied as at least one processor and/or may comprise at least one processor, wherein the processor may be configured, specifically by software programming, for performing one or more operations. In some embodiments, the controller may be suitable for processing the plurality of spectral data sets, in particular for converting analog data into digital data. Analog data may comprise for example electrical signals comprising electrical currents. Analog data may be generated by a detector. Digital data may comprise data represented in a sequence of bits. Digital data may be suitable for processing, e.g. by a processor and/or transferal to a device by means of wired and/or wireless connection.

A processor is a processor comprising a central processing unit (CPU) and/or a graphics fit units (GPU) and/or an application specific integrated circuit (ASIC) and/or a tensor processing unit (TPU) and/or a field-programmable gate array (FPGA). The processor may also be an interface to a remote computer system such as a cloud service. The processor may include or may be a secure enclave processor (SEP). An SEP may be a secure circuit configured for processing the spectra. A "secure circuit" is a circuit that protects an isolated, internal resource from being directly accessed by an external circuit. The processor may be an image signal processor (ISP) and may include circuitry suitable for processing images, in particular.

Aggregating a plurality of spectra as used herein is a broad term and is to be given its ordinary and customary meaning to a person of ordinary skill in the art and is not to be limited to a special or customized meaning. Aggregating a plurality of spectra may refer to processing the plurality of spectra into at least one spectrum. In particular, aggregating a plurality of spectra may comprise merging the information of at least two spectra into one spectrum. The number of aggregated spectra may be smaller than the number of spectra included in the plurality of spectra. Aggregating may include performing mathematical operations to the plurality of spectra. For example, aggregating a plurality of spectra may include summing the intensities and/or the measures related to the intensity, in particular the intensities and/or measures related to the intensity may correspond to the same measure for the energy of light. Furthermore, aggregating may include dividing the sum over the intensities and/or the measures related to the intensity by the number of spectra comprised in the plurality of spectra. Another example for aggregating a plurality of spectra into at least one aggregated spectrum may include using Li median. Li median may be used as described in “The multivariate Li-median and associated data depth” by Y. Vardi et al. (2000) PNAS 97 (4) 1423-1426. Plurality of spectra may be normalized before or after being aggregated. Another example may include multiplying the intensities and/or the measures related to the intensity with a weighting factor before adding the resulting intensities and/or the measures related to the intensity. In some embodiments, aggregating a plurality of spectra may include aggregating spectra more than once, e.g. by aggregating at least two spectra into one spectrum followed by aggregating the resulting spectrum from the first aggregation with another spectrum, in particular another aggregated spectrum.

A “computer program” includes instructions for executing the steps of the method according to the present invention in one or more of the embodiments enclosed herein, in particular when the program is executed on a computer or computer network. Specifically, the computer program may be stored on a computer-readable data medium. As used herein, the terms “computer-readable data medium” specifically may refer to non-transitory data storage means, such as a hardware storage medium having stored thereon computer-executable instructions. The computer-readable data medium specifically may be or may comprise a storage medium such as a random-access memory (RAM) and/or a read-only memory (ROM). Thus, specifically, one, more than one or even all of method steps as indicated herein may be performed by using a computer or a computer network, preferably by using a computer program. Further computer program may have program code means, in order to perform the method according to the present invention in one or more of the embodiments enclosed herein, in particular when the program is executed on a computer or computer network. Specifically, the program code means may be stored on a computer-readable data medium. Further disclosed and proposed herein may be a data medium having a data structure stored thereon, which, after loading into a computer or computer network, such as into a working memory or main memory of the computer or computer network, may execute the method according to one or more of the embodiments disclosed herein. As used herein, a computer program refers to the program as a tradable product. The product may generally exist in an arbitrary format, such as in a paper format, or on a computer-readable data medium. Specifically, the computer program may be distributed over a data network. The term “non-transitory” has the meaning that the purpose of the data storage medium is to store the computer program permanently, in particular without requiring permanent power supply. „lnput“ and/or “output” comprises of one or more of serial or parallel interfaces or ports, USB, Centronics Port, FireWire, HDMI, Ethernet, Bluetooth, RFID, Wi-Fi, USART, or SPI, or analogue interfaces or ports such as one or more of ADCs or DACs, or standardized interfaces or ports to further devices.

“Material information” refers to quantitative information and/or qualitative information and/or material properties. Quantitative information may comprise information associated with an amount of at least one chemical substance in a material. The amount may be a relative amount relating the amount of the at least one chemical substance to the total amount of the material and/or to the amount of at least another chemical substance. The amount may be an absolute amount of a chemical substance in a material. Amount may be determined as amount-of-substance fraction, mass fraction volume fraction and/or the like. Qualitative information may comprise to information suitable for identifying the at least one chemical substance comprised in a material. A chemical substance may be identified via at least one structural part of the chemical substance. Structural part may correspond to at least one atom comprised in the chemical substance. Preferably, structural part may correspond to at least two atoms comprised in the chemical substance, most preferably the two atoms may be connected via a chemical bond. For example, a structural part may comprise a chemical functional group. Furthermore, qualitative information and/or quantitative information may be related to material properties. Material properties may comprise physical and/or chemical properties. A physical property may refer to properties describing the physical state of a material. Physical property may be one of the following: mechanical properties, electrical properties, optical properties, thermal properties or the like. Examples for physical properties may be concentration, color or absorption. A chemical property may be a property defined by the structure of the at least one chemical substance. Chemical property is a property that can be established only by changing the structure of the at least one chemical substance. Examples for chemical properties may be acidity, oxidation state or reactivity.

In some embodiments, the spectrometer device may be portable, in particular handheld. The spectrometer may be readily applied, in particular by non-expert users thus, enabling a simple and robust utilization of spectroscopic methods. Said methods may be even applied in field and far off from laboratory conditions opening spectroscopic methods for everyday applications and especially for dynamic settings.

In some embodiments, the spectrometer device may be a part of a system for generating a spectrum of a sample comprising an input configured to receive a plurality of spectral data sets suitable for obtaining a plurality of spectra from the spectrometer device according to the method with all embodiments as described herein, a processor configured to aggregate at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra, an output configured to provide the at least one aggregated spectrum. In some embodiments, a mobile device such as a smartphone may comprise the input, the processor and/or the output. In other embodiments, a cloud system may comprise the input, the processor and/or the output. In particular, the mobile device may be suitable for receiving the plurality of spectral data sets suitable for obtaining a plurality of spectra and for transferring the plurality of spectral data sets to a cloud system. In some embodiments, the mobile device may be further suitable for obtaining a plurality of spectra from the plurality of spectral data sets. The cloud system may be suitable for aggregating the plurality of spectra into at least one aggregated spectrum and providing the at least one aggregated spectrum, e.g. to a user device.

Specifically, the detector and the controller are comprised in a spectrometer device suitable for generating and/or providing a plurality of spectral data sets and/or wherein the processor is comprised in a mobile device and/or a cloud system.

In some embodiments, the spectrometer device may comprise an optical element. An optical element is configured for changing the intensity of at least a part of light with a defined wavelength at one spatial point. The intensity of at least a part of light with a defined wavelength at one spatial point may be changed by guiding light spatially and/or changing properties of light. An optical element may be suitable for guiding light spatially and/or changing properties of light. Guiding may include but is not limited to changing direction of light, focusing, defocusing or the like. Properties of light may comprise energy, phase, intensity, power or the like. Properties of light may be changed by interaction of light with matter and/or light. An example for light-light interaction may be interference. Examples for light-matter interaction may comprise absorption, scattering, diffraction, refraction orthe like. Optical element may comprise at least one of the following: a mirror, a beam splitter, a dispersive element such as a prism, an interferometer such as a Michelson interferometer, a diffractive element, a slit or any combination thereof.

In some embodiments, the spectrometer device may further comprise the processor configured to receive a plurality of spectra from the controller and to aggregating at least a part of the plurality of spectra into at least one aggregated spectrum in response to having determined at least one difference among the plurality of spectra. By doing so, the spectrometer can directly provide the at least one aggregated spectrum and/or results derived from the at least one aggregated spectrum. This enables a shortening of the time needed to analyze the spectral data sets after having generated the data sets and on-the-go analysis of just now generated data sets since no connection to an external computing device needs to be established in order to process the plurality of spectral data sets. Additionally, on-the-go analysis fosters remeasuring of spectral data sets where it is useful and necessary and saves capacity for saving more aggregated spectra than compared to the higher number of non-aggregated spectra. In particular memory resources are used more efficiently since no capacity needs to be wasted for low-quality spectra. At the same time, low-quality spectra are replaced by high-quality spectral data sets ultimately resulting in an improved data processing for the user..

In some embodiments, the plurality of spectral data sets is generated by means of at least one of absorption spectroscopy, emission spectroscopy, analysis of elastic scattering and/or inelastic scattering, reflection spectroscopy, impedance spectroscopy, resonance spectroscopy, nuclear spectroscopy. In particular, I R spectroscopical methods such as NIR or MIR, UWis spectroscopy or fluorescence spectroscopical methods may be suitable for generating a plurality of spectral data sets since these methods may be easy to apply also apart from more ideal laboratory conditions.

In some embodiments, the plurality of spectral data sets may be received from a spectrometer device, in particular a spectrometer device as described herein.

In some embodiments, spectral data set comprises at least one of a spectrum or an interferogram. In particular, a plurality of spectral data sets may comprise at least one spectral data set obtained from an infrared absorption measurement.

In some embodiments, the user may be provided with information related to the at least one difference and/or is invited to provide input associated with aggregating the plurality of spectra into at least one aggregated spectrum. Information related to the at least one difference may comprise sample information, measurement information, user information or the like. Sample information may be information associated with the sample. Examples may regard at least one of the spatial expansion, quantitative information, qualitative information, texture or the like. Measurement information may be information associated with generating at least one spectral data set. Measurement information may comprise temporal, spatial information, information regarding the surrounding (metadata) conditions or the like associated with generating of at least one spectral data set of the plurality of spectral data set. Examples may be a temperature, air pressure, point in time, location during measurement or similar measurement information. User information may be information associated with the user generating at least one spectral data set. User information may comprise user guidance for generating at least one spectral data set, account information such as a warning caused due to an unauthorized user generating spectral data sets or the like. In an example, user guidance may advise the user to repeat a measurement with a smaller distance between the spectrometer device and the sample and/or change conditions of the surrounding such as regulating room temperature, aerating the room or move a body part away from the measurement setup, e.g. to avoid disturbance of the measurement by the user. By doing so, the aggregated spectrum is customizable and may provide high-quality results. Additionally, errors due to user mishandling may be avoided and accuracy of measurement may be increased. The user may provide input via a graphical user interface (GUI), e.g. GUI of an application. The user may provide input via an application, e.g. a web application and/or application on a mobile device. The input of the user may comprise approval, validation and/or conformation, e.g. regarding a suggestion of the application. The input may further comprise, for example, a suggestion selected by the user. In some embodiments, determining the at least one difference is further based on user input. User input may be used to determine the at least one difference, in particular for selecting the at least one difference. User input may comprise a value for a usual deviation in a measurement. Thus, the user may input a value, e.g. a threshold value below which no difference may be determined and above or equal to the threshold value a difference may be determined among the plurality of spectra. For this purpose, the user may input a numerical value. In other embodiments, the user input may comprise approval, validation and/or conformation based on which at least a part of the plurality of spectra may be aggregated. Hence, a further step of validating, approving and/or conforming aggregating at least a part of the plurality of spectra, in particular by a user, may be included in the method disclosed herein.

In certain embodiments, the information related to the at least one difference comprises a representation of the plurality of spectral data sets. The representation may be a graphical representation, e.g. by displaying spectra. Spectra may be displayed with a display device. An application may be suitable for displaying spectra.

In some embodiments, the at least one difference may be determined based on a comparison among the plurality of spectra. From a comparison, a feature, in particular a distinguishing feature, may be derived. In some embodiments, the feature may be comprised in only a part of the plurality of spectral data sets. Comparing may include subtracting the intensities and/or measure derived from the intensity of two spectral data sets, in particular corresponding to the same measure for the energy of light. Subtracting may reveal a value other than zero when a difference may be determined. In some embodiments, the value determined by subtracting may be unequal to zero and may be compared to a threshold. Such a value may exceed the threshold and thus, a difference, in particular a feature, may be determined. In some embodiments, the value may not exceed the threshold and thus, no difference, in particular no feature, may be determined. A difference may be determined for example among spectra belonging to two different kinds of chemical species and/or between high-quality and low-quality spectra.

A spectrum may be classified as low-quality and/or high-quality by the user, the controller, and/or classification model. The classification model may be a data-driven, deterministic or hybrid model. A data-driven classification model may comprise at least one machine-learning architecture and model parameters. For example, the machine-learning architecture may be or may comprise one or more of: linear regression, logistic regression, random forest, piecewise linear, nonlinear classifiers, support vector machines, naive Bayes classifications, nearest neighbors, neural networks, convolutional neural networks, generative adversarial networks, support vector machines, or gradient boosting algorithms or the like. In the case of a neural network, the model can be a multi-scale neural network or a recurrent neural network (RNN) such as, but not limited to, a gated recurrent unit (GRU) recurrent neural network or a long short-term memory (LSTM) recurrent neural network. Forthis purpose, a model may be trained. The term “training”, also denoted learning, as used herein, is a broad term and is to be given its ordinary and customary meaning to a person of ordinary skill in the art and is not to be limited to a special or customized meaning. The term specifically may refer, with-out limitation, to a process of building the classification model, in particular determining and/or updating parameters of the classification model. The classification model may be at least partially data-driven. For example, the classification model may be based on experimental data, such as data determined by illuminating a plurality of living organisms such as humans and recording the reflection images. For example, the training may comprise using at least one training dataset, wherein the training data set comprises reflection images, e.g. of a plurality of humans with known condition measures. For example, if the neural network is a feedforward neural network such as a CNN, a backpropagation-algorithm may be applied for training the neural network. In case of a RNN, a gradient descent algorithm or a backpropagation-through- time algorithm may be employed for training purposes. A deterministic classification model may comprise an algorithm including instructions for determining the at least one difference. The algorithm may implement a definition for a difference. Based on this definition, the model may determine at least one feature, in particular the model may group the spectra. A deterministic model may comprise a definition based on a numerical value for a usual deviation. Further, the deterministic model may comprise instructions for comparing the plurality of spectra, e.g. by comparing two spectra at a time over the whole spectral range or by comparing all spectra at the same measure for the energy of light. A hybrid classification model may comprise of a data-driven model with deterministic constraints. For example, a constraint may include limitations of certain values.

In some embodiments, the at least one difference is determined based on a feature being comprised in only a part of the plurality of spectral data sets. This is advantageous in the sense that one can account for measurement uncertainty of spectroscopic methods by introducing the threshold. Data is not unnecessarily discarded or left out from analysis due to a non-significant deviation. Ultimately, the processing of spectra is improved. The threshold can be selected based on the required certainty that a difference corresponds to a feature associated with a difference in material information and is not caused by measurement uncertainty, so minimizing the false positive rate. This comes at the cost of identifying too many differences as associated with a difference in material information, i.e. yield a high false negative rate. The threshold is hence usually a compromise between minimizing the false positives rate and keeping the false negative rate at a moderate level. The threshold may be selected to obtain an equal or close to equal false negative rate and false negative rate. The threshold may be a numerical value.

Spectra may be grouped prior to being aggregated. In some embodiments, the spectra comprising the feature may be grouped and/or the spectra not comprising the feature may be grouped. No difference may be determined among the spectra in a group, in particular spectra in a group may comprise the at least one same feature. Followingly, a group may comprise at least one spectrum. Spectra in a group may be tagged and/or corresponding spectral data sets may be tagged, in particular with the same tag. Tagging data sets may enable a readily availability of groups of spectra, thereby increasing the transparency of linking spectra. Further, linked spectra may increase the analysis of spectra if analysis of single spectra is required, e.g. to gain a deeper understanding of a sample consisting of several parts with different chemical composition. The plurality of spectra may be represented via a graphical user interface (GUI), in particular a group of spectra may be represented via a GUI. For example, grouping the plurality of spectra may result in at least two groups, wherein at least one group may comprise spectra classified as high- quality and/or at least another group may comprise spectra classified as low-quality.

In some embodiments, a user selects the at least one difference among the plurality of spectral data sets. Differences may be determined by an algorithm. Listing of the at least one difference may be provided to the user, for example via a graphical user interface. The user may select at least one difference among the at least one difference provided. The user may be enabled to determine at least one difference among the plurality of spectral data sets, e.g. by an application. By doing so, the aggregated spectrum is customizable and may provide more high-quality results. In particular, if the user is a user advanced in spectroscopical methods, the user may input his knowledge and his desired results.

In some embodiments, at least one spectral data set of the plurality of spectral data sets different from at least another spectral data set of the plurality of spectral data sets is discarded and/or remeasured, in particular spectral data sets, wherein corresponding spectra may be classified low-quality, may be discarded and/or remeasured. Spectral data sets discarded and/or remeasured may be grouped into at least one group. By doing so, memory resources are used more efficiently since no capacity needs to be wasted for low-quality spectral data sets. At the same time, low-quality spectral data sets are replaced by high-quality spectral data sets ultimately resulting in an improved data processing for the user. Specifically, the spectra comprising the feature or the spectra not comprising the feature may be aggregated into at least one aggregated spectrum. In this context, not using spectra for aggregation may be referred to as excluding spectra from aggregation. Excluding spectra from aggregation may further include deleting the data corresponding to the spectra on at least one device. Excluding only a part spectra comprising the feature is advantageous since reducing the amount of data to be processed saves computational resources and energy consumption. Further, excluding spectra may include tagging the corresponding data set. Tagging a data set may be advantageous in a scenario, where the data is analyzed at a later point in time to save time for determining the at least one difference again, In particular, a signal may be output that triggers an additional measurement based on spectra including the feature or spectra not including the feature. Additional measurement may include repeating the measurement. Repeating the measurement may include measuring the same sample with the same measurement device, e.g. spectrometer device.

In some embodiments, the plurality of spectral data sets is provided and/or the aggregated spectra are received, in particular by a user and/or user device.

In some embodiments, the plurality of spectral data sets may comprise at least one IR spectral data set.

Any disclosure and embodiments described herein relate to the methods, the devices, the use, the computer program element, the computer-readable data medium lined out above and vice versa. Advantageously, the benefits provided by any of the embodiments and examples equally apply to all other embodiments and examples and vice versa. The devices may be suitable for carrying out the steps of the methods.

Further possible implementations or alternative solutions of the invention also encompass combinations - that are not explicitly mentioned herein - of features described above or below in regard to the embodiments. The person skilled in the art may also add individual or isolated aspects and features to the most basic form of the invention. It shall be understood that a preferred embodiment of the invention can also be any combination of the dependent claims or above embodiments with the respective independent claim.

Further embodiments, features and advantages of the present invention will become apparent from the subsequent description and dependent claims, taken in conjunction with the accompanying drawings.

Brief Description of the Figures

Figure 1 illustrates a flow diagram of an example embodiment of a method for generating a spectrum of a sample (100).

Figure 2 illustrates a flow diagram of an example embodiment of a method for generating a spectrum of a sample (200).

Figure 3 illustrates a block diagram of an example system and an example spectrometer device for generating a spectrum of a sample (300). Figure 4 illustrates exemplary representations optical spectra (400).

Figure 5 illustrates an exemplary graphical user interface suitable for providing information to a user (500).

In Fig. 1 a flow diagram of an example embodiment of a method for processing optical spectra (100) is illustrated.

In a first step, a plurality of spectral data sets is received (110). The spectral data sets are suitable for obtaining a plurality of spectra. The spectral data sets may be received from a spectrometer having generated the spectral data sets and/or from a data storage. Data storage may comprise a local storage of a device such as computers, servers, memory sticks or the like and/or external storage as provided by a cloud computing environment and/or a decentralized environment. Spectral data sets generated with the spectrometer may be aggregated directly after the measurement series is generated and requirements are sufficient such as battery and internet connection. The spectrometer may be connected in a wired and/or wireless manner to a data storage and/or may establish a connection after having generated a measurement series. In some embodiments, the spectral data sets may be stored locally at the spectrometer and/or may be transmitted at a later point in time, e.g. when several measurement series are measured, the user decides to transfer the data, battery is recharged, wired and/or wireless connection is established, spectrometer’s internal storage has reached or is close to reaching a certain occupancy or the like. The plurality of spectral data sets may be received together in shared data packages, by means of one data transferal task and/or separately, e.g. data set by data set or even a spectral data sets may be transferred in parts. In some embodiments, spectral data sets may comprise a normalized spectrum. Normalization may be performed after having generating and/or after having transferred the spectral data sets. A plurality of spectra may be obtained from the plurality of spectral data sets by applying mathematical operations to the plurality of spectral data sets and/or the plurality of spectral data sets may comprise a plurality of spectra. In some embodiments, the plurality of spectral data sets may comprise at least one spectrum and at least one prestage of a spectrum. In this exemplary scenario, the at least one spectrum may be processed without performing a mathematical operation and/or the at least one prestage of a spectrum may be transformed into a spectrum performing a mathematical operation. For example, the plurality of spectral data sets may comprise at least one interferogram which may be transformed into a spectrum by means of Fourier transform.

In a next step, at least a part of the plurality of spectra is aggregated into at least one aggregated spectrum based on at least one difference among the spectra (120). Forthis purpose, a difference may be determined. This may be facilitated by means of an algorithm. The algorithm may be suitable for comparing and/or classifying the plurality of spectra. In other embodiments, the algorithm may implement a data-driven model trained to group spectra based on at least one difference. Such a data-driven model may be a classifier. The algorithm may be programmed with instructions defining/specifying/presetting the at least one difference among the plurality of optical spectra. The at least one difference may be described by what the at least one difference may be (e.g. position of peaks, shape of peaks, intensity/height of peaks) and a threshold to differentiate between sufficient differences and non-sufficient differences. Non-sufficient differences may refer to differences among the spectra that are caused by measurement errors and/or differences within a sample intended to be measured together. In an example, a sample may be inhomogeneous (e.g. a polymer blend) and the user may desire to determine a mean value of a measure and/or a ratio. Hence, the requirements for a difference may be use-case specific. Sufficient differences may refer to differences among the spectra where the user may not desire to aggregate the spectra. Reasons for this may be that the user does not want to determine a mean value of a measure, but may desire to analyze a certain part, e.g. the sample may be placed on a table and some spectra may comprise information of only the desired sample and some spectra may comprise information of the sample and the table. In the before mentioned example, the spectral information of the table may be undesired and thus, spectra comprising spectral information of the table should not be aggregated with the spectra not comprising any spectral information of the table.

When a plurality of spectra comprises at least one difference, then the spectra may differ in a way that at least one spectrum may comprise a feature that the rest of the plurality of spectra may not comprise. Based on the at least one difference, the spectra for the aggregation may be selected. The spectra and/or the at least one difference may be selected by a user and/or an algorithm as described above. Spectra comprising a feature identified as a difference among the plurality of spectra may be grouped. In a group of spectra, only those spectra may be included which include the at least one common feature. Spectra in a group may be aggregated. Aggregating spectra may be used to yield one aggregated spectrum from a plurality of spectra. This can be facilitated by adding the intensities of the spectra of the same measure for the energy of the light, e.g. wavelength, frequency, wavenumber or the like. In some embodiments, the resulting intensity may be divided by the number of intensities added. The aggregated spectrum may be a normalized or an unnormalized spectrum. Unnormalized spectrum may be a result of aggregating unnormalized spectra. Normalized aggregated spectrum may be a result of aggregating unnormalized spectra and then normalizing the yielded aggregated spectrum or aggregating normalized spectra and dividing the intensity by the number of spectra used for aggregation. In some embodiments, only a part of spectra may be aggregated.

In some embodiments, the plurality of spectra used for aggregating may be normalized spectra.

In other embodiments, the plurality of spectra used for aggregating may not be normalized spectra. Thus, spectra may be normalized before or after aggregation. This can be done by normalizing the plurality of spectra first and then aggregating the spectra or aggregating the unnormalized spectra first and then normalizing the at least one aggregated spectrum.

The aggregated spectrum is provided (130). The aggregated spectrum may be provided to a user by means of a user device (e.g. smartphone, computer, laptop etc.) and/or a user account (e.g. user account to access a data storage such as cloud storage or an account to request and/or receive the data). User device and/or user account may be accessed by the user. User device and/or user account may be suitable for downloading, displaying, sharing, processing, generating, and/or storing spectra.

In an example, a user may have initiated the generation of the plurality of spectra with a spectrometer, in particular a handheld spectrometer. The generated spectra may be transferred to a cloud system and/or a mobile device such as a smartphone. The cloud system and/or a mobile device may in addition be configured to receive and/or store the spectra. Additionally or alternatively, the cloud system and/or the mobile device be suitable for providing the spectra and/or processing the spectra. After having generated the spectra, the user may desire to access the generated data. For this purpose, the user may operate an application on a smartphone and/or computer. By using the application, the user may access the spectra from the cloud system, in particular the at least one aggregated spectrum. The application may provide a graphical representation of the spectra, in particular the at least one aggregated spectrum via a GUI. Further, the user may process the spectra at least partially, e.g. select spectra for aggregation, with the application. In some embodiments, a user account may be accessed to download the spectra and/or transfer the spectra.

In Fig. 2 a flow diagram of an example embodiment of a method for generating a spectrum of a sample (200) is illustrated.

A plurality of spectral data sets is received (210) as described in the context of Fig. 1.

A difference may be determined among the plurality of spectra (220). The difference may be selected by a user. The user may select the difference, e.g. from a list. The difference may be determined as described in the context of Fig. 1 . A difference among the plurality of optical spectral data sets when a feature may be present. Such a feature may be a distinguishing feature. Distinguishing feature may be a feature comprised in at least one of the plurality of spectra and may not be comprised in at least another one of the plurality of spectra. Examples for a feature may be an additional peak and/or a shifted peak. In some embodiments, determining the at least one difference is further based on user input. The user may be invited to provide input associated with aggregating the plurality of spectra. For this purpose, the user may select the at least one difference. The at least one difference may be selected and/or identified by the user. The user may be provided with a list comprising suggestions for example. The user may select the at least one difference by using an application. The application may offer a suggestion to which the user may agree. The application may visualize the spectral data sets, e.g. as spectra. By visualizing the plurality of spectral data sets, the user may be enabled to select at least one difference, e.g. by selecting at least one feature comprised in the corresponding at least one of the spectra used for visualization. The difference may be determined by performing mathematical operations to data comprised in the spectral data sets. For example, a difference may be a difference in intensity which may be calculated by subtracting intensity values. Another example may be a difference in peak position which may be determined by subtracting wavelengths or any other value of a measure related to the energy of the light used in when generating spectral data sets. In some embodiments, the user input may be associated with a validation of aggregating of spectra. Hence, aggregating at least a part of the plurality of spectra may be validated and/or conformed based on user input (230).

At least one of the plurality of spectra may be discarded and/or remeasured (240). Reasons for discarding and/or remeasuring a spectrum may be for example measurements of undesired parts of a sample and/or spectra with too high signal-to-noise-ratios. Discarding a spectrum may refer to deleting the data related to the spectrum and/or excluding the data related to the spectrum from aggregation. The data may be removed locally from at least one device and/or from a plurality of devices storing at least a part of data related to the spectrum to be discarded. In some embodiments, discarding a spectrum may result in generating a new spectrum and/or at least one additional spectral data set may be generated. This can involve instructions for the user on how to generate the new spectrum. Depending on the information related to the at least one difference, a source of error may be identified. For reasons of prevention, instruction relating to the source of error of generating the spectrum discarded or to be discarded may be generated and provided to the user. Discarding at least one of the plurality of spectra may be preceded by an inquiry to the user, e.g. by inviting the user to provide input. The user may verify the operation (230). This is advantageous since deleted data may not restored and is lost for further analysis. By asking the user for input, e.g. in the form of confirmation, unintended discarding may be prevented.

At least a part of the plurality of optical spectra is aggregated into at least one aggregated spectrum (250) as described in the context of Fig. 1.

The at least one aggregated optical spectrum is provided (260) as described in the context of Fig. 1. In Fig. 3a, a block diagram of an example spectrometer device for generating a spectrum of a sample (300) is illustrated. The spectrometer device may comprise a detector (310) and a controller (320).

The detector (310) is configured to generate a plurality of spectral data sets. The detector (310) may detect light. The light may have interacted with a sample and/or the light may have passed through the sample and/or may have been reflected by a sample before it may be received by the detector (310). In response to illuminating the detector (310) with electromagnetic radiation, a signal may be generated by the detector (310). The signal may be suitable for determining an intensity of the incident electromagnetic radiation. With this information and a measure for the energy of the electromagnetic radiation, a plurality of spectral data sets may be generated. The measure for the energy of electromagnetic radiation may be determined and/or received by the detector (310).

The controller (320) is configured to provide the plurality of spectral data sets to a processor configured to aggregate at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra. The controller (320) may be operatively coupled to the detector (310) and/or the processor (330). Controller may be suitable for receiving a plurality of spectral data sets, e.g. from the detector (310). In some scenarios, data generated by the detector (310) may comprise raw data and/or analog data. Such data may be converted by the controller (320) such that the data may be suitable for obtaining a plurality of spectra. For example, converting data may include converting analog data into digital data. This may be facilitated e.g. with an analog-to-digital converter that may be part of the controller (320). In some scenarios, the controller (320) may be suitable for transferring the plurality of spectral data sets generated by the detector (310) to the processor (330). Further, the controller (320) may comprise the processor (330) and/or may be connected to the processor (330). Controller (320) may be suitable for controlling the processor (330) and/or the detector (310).

The processor (330) may be configured to aggregate at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra. The processor may be independent of the spectrometer device. Such a processor (330) may be comprised in a decentralized computational network such as a cloud infrastructure and/or in a local computer and/or in a mobile device. In particular, the processor (330) may not be a part of the spectrometer device (300), e.g. the processor (330) may be part of a system for generating a spectrum of a sample as described in the context of Fig. 3b. In some embodiments, the processor may be part of the spectrometer device (300). In some embodiments, the processor may be a part of the controller (320), wherein the controller (320) may be configured to aggregate at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra. An example for such a controller may be a microcontroller or a system on a chip. Such a microcontroller may comprise a processor (e.g. CPU), a memory, an input and/or an output. In some embodiments, the controller may transform an interferogram into a spectrum. In some embodiments, the spectrometer device may further comprise an optical element. The optical element may be configured for changing the intensity of at least a part of light with a defined wavelength at one spatial point. For this purpose, the optical element may comprise a material with indices of refraction different for different wavelengths, a so-called dispersive element. For example, certain glass types show a larger index of refraction for violet light than for red light, thereby eventually causing violet light to travel slower through the material. As a consequence, the light may be dispersed during travelling through the material. By leading the dispersed light with mirrors and/or slits, light with a defined wavelength may be selected. The selected part of the light may be used to illuminate the sample. Another option for changing the intensity of at least a part of light with a defined wavelength at one spatial point may be the use of interferometers such as the Michelson interferometer. An interferometer may be an apparatus comprising a beam splitter and mirrors. Incoming light may be split into at least two beams and may be guided such that interference of the at least two parts of the split beam may occur. Thereby, at least one part of the split beam may be guided such that its pathway between the beam splitter and the point where interference may occur may be varied. As a consequence, the interference pattern may be varied by varying the pathway. Depending on the characteristic of the at least two parts of the light (wavelength and phase), the interference pattern may comprise different intensities for different wavelengths comprised in the light. By applying a Fourier transform, the interferogram may be converted into a spectrum. Thus, interferograms may be an example for spectral data. The interference pattern may be projected onto the sample. The sample may (e.g. by reflecting, absorbing, scattering or the like) or may not interact with the light. As a result, the intensity of the light after interacting or not interacting may be measured with a detector (320). It is to be understood that emitted light from the sample in response to light interaction may also be determined and/or used to generate a spectral data set. Further, it is to be understood that generating a spectrum from a sample may include processing spectral data sets.

In Fig. 3b an exemplary embodiment of a system for generating a spectrum of a sample (300b) is illustrated.

The system (300b) may comprise a spectrometer device, e.g. as described in the context of Fig. 3a, and a mobile device and/or a cloud system. The spectrometer device may further comprise a radiation source. The sample may be illuminated with light. The light may be emitted from a radiation source and may be allowed to interact with the sample. At least a part of the light may be transmitted to a photodetector. The photodetector may be suitable for generating a signal, e.g. an electrical signal, in response to being illuminated with light. The photodetector may be suitable for detecting light with a specific energy. For example, the photodetector may be configured to detect light e.g. in the infrared, visible or ultraviolet range. Detecting light may refer to generating a signal indicating an intensity of the incident light. By determining the intensity from the signal with the corresponding measure of the energy of the light a spectral data set may be generated. In some embodiments light of several energy values may be generated and allowed to interact with the sample. A corresponding interferogram relating a measure for the intensity of the light to a measure of the phase difference of at least a part of the light may be a result of detecting light with a photodetector. An interferogram may be transformed into a spectrum, e.g. by means of a Fourier transform. This may be facilitated with a processor and/or controller. In a scenario, where spectral data set comprises only spectra, a transformation is not necessary and the spectra may be directly used for aggregation. In some embodiments, controller may be or may comprise a processor suitable for aggregating a plurality of spectra. In other scenarios, controller may be suitable for controlling a processor suitable for aggregating a plurality of spectra. For the latter purpose, the controller may be communicatively and/or operatively coupled to the processor. In some embodiments, the processor may be part of a device, in particular a display device. The display device may be suitable for displaying at least one spectrum and/or operating an application via a user interface. In certain scenarios, the device may be connected to a cloud system. Such a cloud system may comprise at least one processor and may be suitable for aggregating a plurality of spectra. In some scenarios, the device may be used for transmitting the spectra to the cloud system. Further, the cloud system may be configured to store at least one spectral data set and/or distributing the spectral data set by means of wireless communication.

In Fig. 4 exemplary representations of spectral data sets (400) are illustrated. Spectra may be a result of different kind of spectroscopy such as IR spectroscopy, in particular MIR, NIR and/or FTIR spectroscopy, UWis spectroscopy, fluorescence spectroscopy and raman spectroscopy. Hence, the spectra may be represented with different units and/or measures on the axes. Usually, in spectroscopical methods, a measure associated with light intensity or the change in light intensity related to light absorption, reflection, scattering and/or emission is plotted against a measure for the energy of the light. Measures associated with light intensity or the change in light intensity may be refer to the intensity of the light, the change in intensity of the light, the absorption, the extinction and/or any other measure relating at least one of the measures. As an example, depending on the kind of spectroscopy, the measure for the energy of the light may be wavelength e.g. given in m, energy e.g. given in eV or J, wavenumber e.g. given in nr¹ , frequency e.g. given in s^-1 or the like. In some embodiments, the intensity of absorption, reflection, scattering and/or emission of light at one certain wavelength may be of interest and may be plotted as it can be seen in Fig. 4a. The signal intensity at the certain wavelength may be plotted over time indicating a measurement series over time of a sample. The sample may comprise a substance that may show a characteristic absorption behavior at the wavelength chosen, e.g. the substance may strongly absorb the light of the certain wavelength. In an example, the sample may have a hole in the middle and/or the composition in the middle of the sample may comprise a low amount of the substance showing the characteristic absorption behaviorthan the composition of the outer parts of the sample. Thus, the generating spectral data sets by moving a spectrometer from one outer part of the sample through the middle to another outer part of the sample may result in a decreasing signal intensity for the certain wavelength when approaching the middle and an increasing signal intensity when departing from the middle. Another option may be that the sample may be moved and the spectrometer may be in a fixed position and/or both are moved. In other embodiments the sample may be contaminated and/or may be coated in the middle part such that the signal of the substance with the characteristic absorption behavior is decreased. A corresponding temporal plot may be similar to the one that can be seen in Fig. 4a. In Fig. 4a t₂ may be a point in time where the generation of spectra associated with the middle part starts and t₂ may correspond to a point in time when the spectrometer passed the middle part of the sample. As described herein, an intensity drop may be an example for the at least one difference. To visualize the intensity drop, one may choose the representation as it can be seen in Fig. 4a. By plotting the signal intensities, the time interval with a signal intensity below a threshold may be identified. The threshold may be indicated in the graph as a horizontal line at a specific signal intensity. Depending on the use case, the threshold may be at different values as indicated by the two horizontal lines. For example, the spectrometer may be moved across the sample over time and a first part (until ti) may be interesting with specific characteristics resulting in the first threshold. The middle part of the sample (between ti and t₂) may be not relevant for the measurement since this can be a completely different material and/or may be a hole in the sample. The part of the sample measure after t₂ may be again of interest but may correspond to a part of the sample with a different composition. Criteria for an intensity drop may be specified by the algorithm implementing the methods described herein or may be selected by the user, e.g. on a use-case specific basis.

Other examples for the at least one difference may be visible in Fig. 4b. Six spectra may be generated at six different points in time (ti, t₂, ts, t4, ts, te). As described in the context of Fig. 4a, each spectrum may correspond to another part of the sample. By comparing all spectra, one may find that spectrum 1 at ti is very similar to spectrum 1 at t₂. Followingly, a difference may not be determined among the first two spectra. The third spectrum at t₃ possesses similar or same peak position and peak shape and decreased intensity compared to the first two spectra. Depending on the criteria for the at least one difference, a difference among the first three spectra may be determined. Followingly, the third spectrum may be classified as different or the spectrum may be in an acceptable range depending on the conditions set by the algorithm implementing the method and/or the set by the user. An even more decreased intensity of the peaks with similar or same peak position and shape can be seen in spectrum 4 at t₄. Similar arguments can be named for the decrease as for spectrum 3. Looking now at spectrum 5 at t₅ and spectrum 6 at t₆ reveals a difference in the position of at least one of the peaks and a difference in the peak shape compared to the peaks in the first two spectra. Spectra 5 and 6 possess peaks with a smaller peak width indicating a different chemical nature of the at least one substance comprised in the sample when measuring at t₅ and t₆ than compared to the first two spectra. Consequently, spectra 5 and 6 may correspond to a different part of the sample or a different material than spectra 1 and 2. By interpreting the six spectra, one may conclude that the sample measured comprises at least two regions with different composition, wherein the first region was measured from ti to t₄ eventually characterized with a decreasing concentration of a certain substance A. After crossing a border to region two at ts the concentration of another substance B may increase until te. If the right peak of spectra 1-4 corresponds to the same wavelength or energy of the light as the left peak of spectra 5-6, the intensity evolution overtime may be similar to the one from Fig. 4a with a different time assignment. Instead of ti and t₂ one may think of points in time around t₄.

In Fig. 5 an exemplary graphical user interface suitable for providing information to a user (500) is illustrated. Such a GUI (500) may be displayed while the user may be operating an application, e.g. an application for analyzing spectra. The GUI (500) may comprise of a header (510) specifying the plurality of spectral data sets, a representation of the spectral data sets (520), options regarding the spectra (530) with recommendations by the application highlighted by underlining and an additional option for the user to select the at least one difference (540) by clicking a button to regroup the spectra. The plurality of spectral data sets may comprise a measurement series overtime. The plurality of spectral data sets may be processed as described in the context of Fig. 1-3. As described in the context of Fig. 4a and 4b, the sample may be moved in relation to the spectrometer while generating the spectral data sets. By doing so, the generated measurement series may be the measurement series as already described in the context of Fig. 4b. The series may be titled “Measurement - wheat”. The title may be chosen by the user. The location of the measurement may be chosen by the user and/or may be determined via GPS. The spectral data sets may be generated by the user and/or initiated by the user in order to determine the water content of the wheat. Therefore, the user may have indicated a target in order to let the application carry out the analysis. The application may determine at least one difference among the plurality of spectra. As discussed in the context of Fig. 4b, the first three spectra may be grouped together since the intensity difference may not have been considered disturbing to a high-quality analysis. The grouping may be indicated by the box surrounding the representation of the three spectra with corresponding options (530) such as analyze, discard and remeasure. The application may have determined a preferred option, herein highlighted by underlining. The user may be given the choice to follow the suggestion of the application or to decide by on his own. The user may select the at least one difference by using the regrouping option (540) or choosing the between the options already given by the application. In response to the user identifying the at least one difference among the spectra, the application may aggregate the plurality of spectra into at least one aggregated spectrum. Based on the at least one aggregated spectrum, the application may carry out the analysis of the spectra and/or send a request indicating the analysis of the spectra. As used herein “determining" also includes “initiating or causing to determine", “generating" also includes ..initiating or causing to generate" and “providing” also includes “initiating or causing to determine, generate, select, send or receive”. “Initiating or causing to perform an action” includes any processing signal that triggers a computing device to perform the respective action.

Claims

1 . A computer-implemented method for generating a spectrum of a sample comprising: receiving a plurality of spectral data sets suitable for obtaining a plurality of spectra, aggregating at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra, and providing the at least one aggregated spectrum.

2. The method according to any one of the preceding claims, wherein the plurality of spectral data sets is generated by means of one of absorption spectroscopy, emission spectroscopy, analysis of elastic scattering and/or inelastic scattering, reflection spectroscopy, impedance spectroscopy, resonance spectroscopy, nuclear spectroscopy.

3. The method according to any of the preceding claims, wherein the spectral data set comprises at least one of a spectrum or an interferogram.

4. The method according to any one of the preceding claims, wherein the at least one difference is determined based on a comparison among the plurality of spectral data sets.

5. The method according to any one of the preceding claims, wherein the at least one difference among the plurality of spectra is determined based a feature included in only a part of the plurality of spectra.

6. The method according to claim 5, wherein the spectra comprising the feature are grouped and/or the spectra not comprising the feature are grouped.

7. The method according to any one of claims 5 or 6, wherein the spectra comprising the feature or the spectra not comprising the feature are aggregated into at least one aggregated spectrum

8. The method according to any one of claim 5-7, wherein a signal is output that triggers an additional measurement based on spectra including the feature or spectra not including the feature.

9. The method according to any one of the preceding claims, wherein the plurality of spectral data sets comprises at least one IR spectral data set.

10. The method according to any one of the preceding claims, wherein determining the at least one difference is further based on user input.

11. A spectrometer device comprising: a detector configured to generate a plurality of spectral data sets, a controller configured to provide the plurality of spectral data sets to a processor for carrying out the steps according to any one of the methods of claims 1-10.

12. The spectrometer device according to claim 11 , wherein the spectrometer device is a part of a system for generating a spectrum of a sample and wherein the spectrometer device is adapted to be operatively coupled to a mobile device and/or cloud system comprising: an input configured to receive a plurality of spectral data sets suitable for obtaining a plurality of spectra from the spectrometer device, a processor configured to aggregate at least a part of the plurality of spectra into at least one aggregated spectrum based on at least one difference among the plurality of spectra, an output configured to provide the at least one aggregated spectrum.

13. A computer program including instructions for executing steps of the method according to claims 1-10.

14. A non-transitory computer-readable data medium storing a computer program according to claim 13.

15. Use of at least one aggregated spectrum obtained by the method according to claims 1-10 for determining material information.