[go: up one dir, main page]

WO2023057523A1 - Procédés et appareils d'étalonnage de capteur - Google Patents

Procédés et appareils d'étalonnage de capteur Download PDF

Info

Publication number
WO2023057523A1
WO2023057523A1 PCT/EP2022/077718 EP2022077718W WO2023057523A1 WO 2023057523 A1 WO2023057523 A1 WO 2023057523A1 EP 2022077718 W EP2022077718 W EP 2022077718W WO 2023057523 A1 WO2023057523 A1 WO 2023057523A1
Authority
WO
WIPO (PCT)
Prior art keywords
target
spectra
comparison
distribution
samples
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/EP2022/077718
Other languages
English (en)
Inventor
Casimir Wierzynski
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Rockley Photonics Ltd
Original Assignee
Rockley Photonics Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US17/886,409 external-priority patent/US12066327B2/en
Application filed by Rockley Photonics Ltd filed Critical Rockley Photonics Ltd
Priority to CN202280067395.XA priority Critical patent/CN118056115A/zh
Priority to EP22798136.2A priority patent/EP4413338A1/fr
Publication of WO2023057523A1 publication Critical patent/WO2023057523A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28—Investigating the spectrum
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28—Investigating the spectrum
    • G01J3/42—Absorption spectrometry; Double beam spectrometry; Flicker spectrometry; Reflection spectrometry
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
    • G01N21/27—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands using photo-electric detection ; circuits for computing concentration
    • G01N21/274—Calibration, base line adjustment, drift correction
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28—Investigating the spectrum
    • G01J2003/283—Investigating the spectrum computer-interfaced
    • G01J2003/2833—Investigating the spectrum computer-interfaced and memorised spectra collection
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2201/00—Features of devices classified in G01N21/00
    • G01N2201/12—Circuits of general importance; Signal processing
    • G01N2201/129—Using chemometrical methods
    • G01N2201/1296—Using chemometrical methods using neural networks

Definitions

  • the present disclosure generally relates to methods and apparatuses for calibrating sensors.
  • Spectrometer sensors configured to measure absorbances as a function of wavelength may be miscalibrated in the sense that an expected wavelength that it is measuring an absorbance for may be different from the actual wavelength that it is the measuring absorbance for. This mis-calibration may arise due to, for example, a manufacturing defect in the spectrometer sensor or deterioration or other changes within the spectrometer sensor during its lifetime. There is therefore a continual need to improve methods and apparatuses for calibrating spectrometer sensors.
  • a method for determining a calibration function includes: calculating a first distance, between a distribution of target spectra and a comparison distribution of spectra; calibrating the distribution of target spectra with a first preliminary calibration function to form a first distribution of calibrated target spectra; calculating a second distance, between the first distribution of calibrated target spectra and the comparison distribution of spectra; determining that the second distance is less than the first distance; and setting the calibration function equal to the first preliminary calibration function.
  • the distribution of target spectra includes one or more spectra measured by a target sensor
  • the comparison distribution of spectra includes one or more comparison distributions corresponding to one or more respective comparison sensors, each of the comparison distributions including one or more spectra measured by the corresponding comparison sensor.
  • the target sensor is an optical spectrophotometer and each of the comparison sensors is an optical spectrophotometer.
  • the comparison distribution of spectra is based on the one or more comparison distributions.
  • the comparison distribution of spectra is the union of the one or more comparison distributions.
  • the target spectra are spectra of at least one target sample, each of the at least one target sample being from a same class of samples
  • the spectra of the comparison distribution of spectra are spectra of at least one comparison sample, each of the at least one comparison samples being from the same class of samples as the at least one target sample is from.
  • the method further includes: illuminating each of the at least one target samples with light of a respective plurality of actual wavelengths of radiation; and measuring, by a target sensor and for each of the at least one target samples, respective portions of the light of the respective plurality of actual wavelengths transmitted through the target sample, wherein the target spectra includes a plurality of absorbance values of the target sample respectively based on the portions of light.
  • the first distance is calculated based on a metric
  • the metric is a Wasserstein distance metric, a Kullback-Leibler divergence metric, a Renyi divergence metric, or an f-divergence metric.
  • the method further includes selecting the first preliminary calibration function from among a family of preliminary calibration functions.
  • the first preliminary calibration function maps a set of nominal wavelengths to a set of calibrated wavelengths.
  • the method further includes defining the first preliminary calibration function by training a neural network to provide a calibrated wavelength of the set of calibrated wavelengths in response to being provided with a nominal wavelength of the set of nominal wavelengths.
  • a processing circuit for determining a calibration function is configured to: calculate a first distance, between a distribution of target spectra and a comparison distribution of spectra; calibrate the distribution of target spectra with a first preliminary calibration function to form a first distribution of calibrated target spectra; calculate a second distance, between the first distribution of calibrated target spectra and the comparison distribution of spectra; determine that the second distance is less than the first distance; and set the calibration function equal to the first preliminary calibration function.
  • the distribution of target spectra includes one or more spectra measured by a target sensor
  • the comparison distribution of spectra includes one or more comparison distributions corresponding to one or more respective comparison sensors, each of the comparison distributions including one or more spectra measured by the corresponding comparison sensor.
  • the target sensor is an optical spectrophotometer and each of the comparison sensors is an optical spectrophotometer.
  • the comparison distribution of spectra is based on the one or more comparison distributions.
  • the processing circuit is in the target sensor.
  • the processing circuit is further configured: to receive the distribution of target spectra; and to receive, from a cloud storage or a memory external to the target sensor, the one or more comparison distributions.
  • the target spectra are spectra of at least one target sample, each of the at least one target sample being from a same class of samples
  • the spectra of the comparison distribution of spectra are spectra of at least one comparison sample, each of the at least one comparison samples being from the same class of samples as the at least one target sample are from.
  • the first distance is calculated based on a metric, and the metric is a Wasserstein distance metric, a Kullback-Leibler divergence metric, a Renyi divergence metric, or an f-divergence metric.
  • the first preliminary calibration function maps a set of nominal wavelengths to a set of calibrated wavelengths.
  • FIG. 1 depicts a method for calibrating a target sensor according to some embodiments.
  • FIG. 2A depicts a range of true absorbance spectra of target samples according to some embodiments.
  • FIG. 2B depicts a target distribution and a secondary comparison distribution according to some embodiments.
  • FIG. 3A depicts a target absorbance function and a comparison absorbance function according to some embodiments.
  • FIG. 3B depicts a calibrated absorbance function and a comparison absorbance function according to some embodiments.
  • processing circuit is used herein to mean any combination of hardware, firmware, and software, employed to process data or digital signals.
  • Processing circuit hardware may include, for example, application specific integrated circuits (ASICs), general purpose or special purpose central processing units (CPUs), digital signal processors (DSPs), graphics processing units (GPUs), and programmable logic devices such as field programmable gate arrays (FPGAs).
  • ASICs application specific integrated circuits
  • CPUs general purpose or special purpose central processing units
  • DSPs digital signal processors
  • GPUs graphics processing units
  • FPGAs programmable logic devices
  • each function is performed either by hardware configured, i.e. , hard-wired, to perform that function, or by more general-purpose hardware, such as a CPU, configured to execute instructions stored in a non-transitory storage medium.
  • a processing circuit may be fabricated on a single printed circuit board (PCB) or distributed over several interconnected PCBs.
  • a processing circuit may contain other processing circuits; for example, a processing circuit may include two processing circuits, an FPGA and a CPU, interconnected on a PCB.
  • a processing circuit may be distributed, e.g., a processing circuit may consist of, or include, a plurality of processing circuits in communication with each other (via wires, fibers, or wireless communication links).
  • a method e.g., an adjustment
  • a first quantity e.g., a first variable
  • a second quantity e.g., a second variable
  • the second quantity is an input to the method or influences the first quantity
  • the second quantity may be an input (e.g., the only input, or one of several inputs) to a function that calculates the first quantity, or the first quantity may be equal to the second quantity, or the first quantity may be the same as (e.g., stored at the same location or locations in memory as) the second quantity.
  • a spectrometer may illuminate a sample with light, and measure light returning to the spectrometer, after propagating through the sample, as a function of wavelength.
  • the ratio of (i) the radiant flux received by the spectrometer to (ii) the radiant flux transmitted by the spectrometer may be referred to, for the sake of brevity, as the “absorbance” even though attenuation mechanisms other than absorption may affect this ratio. This may be accomplished by emitting light at a set of wavelengths and measuring how much power is reflected or transmitted at each wavelength.
  • the sample has a true spectrum of ATrue(A).
  • a common problem is that, because of fabrication variations or deterioration or changes over time, the laser wavelengths are miscalibrated — i.e. , the actual wavelengths emitted and measured will not, in general, match the desired or expected wavelengths. This means that the measured spectrum may be misinterpreted — the absorbance measurements may be associated with the wrong wavelengths.
  • FIG. 1 depicts a method of calibrating a target sensor according to some embodiments.
  • the method includes a first process 100 of building a target distribution Do associated with a target sensor, a second process 200 of comparing the target distribution Do with a plurality of comparison distributions ⁇ Di ... DN ⁇ , wherein N is an integer equal to or greater than 2, and a third process 300 of determining a calibration function that reduces a distance, as measured by a metric, between the target distribution Do and a plurality of comparison distributions ⁇ Di ... DN ⁇ .
  • the “target distribution” may be referred to herein as a “distribution of target spectra.”
  • the first process 100 may include illuminating each of a plurality of target samples with light at a respective plurality of actual wavelengths and measuring, by the target sensor and for each target sample, a plurality of absorbances, of the target sample, respectively associated with the actual wavelengths. If it is uncalibrated, the target sensor may report the plurality of absorbances together with a plurality of respective nominal wavelengths, which are the respective wavelengths that the plurality of actual wavelengths are assumed or expected to be. However, at least some of the actual wavelengths may be different from their respective nominal wavelengths, e.g., due to mis-calibration or fabrication imperfection of the target sensor.
  • one of the target samples may be illuminated by what is believed to be a nominal wavelength, but the target sample may actually be illuminated by an actual wavelength that slightly or significantly differs from the nominal wavelength, e.g., due to a mis-calibration or fabrication imperfection of the target senor.
  • the target samples may all be from the same class of samples.
  • the class of samples may be a type of biological tissue (e.g., muscle, fat, bone, brain tissue, heart tissue, etc.) or a type of biological material (e.g., blood, plasma, bone marrow, etc.).
  • the class of samples may be biological tissue or biological material at a certain region of a person, animal, plant, fungus, etc., such as the abdomen, neck, eye, or wrist of a person.
  • the class of samples may be a type of non-biological material.
  • the target sensor may be utilized to illuminate the target samples.
  • the target sensor may contain one or more lasers.
  • the target sensor may contain one laser configured to generate light of various wavelengths.
  • the target sensor contains a plurality of lasers, each of the plurality of lasers being configured to generate light of a respective wavelength.
  • one or more lasers or light sources utilized to illuminate the target samples may be separate from the target sensor, and the calibration methods and apparatuses described herein may be utilized to calibrate the one or more lasers or light source(s).
  • each measured absorbance value of a sample may be reported with a respective calibrated wavelength that may be closer, to the respective actual wavelength with which the sample was illuminated, than the respective nominal wavelength by which the sample was assumed or expected to be illuminated.
  • each spectrometer measurement may be reported as a pair of vectors, including a vector of wavelengths, and a vector of corresponding absorbances.
  • the pluralities of nominal wavelengths by which the plurality of target samples are respectively assumed to be illuminated may be the same or different.
  • a first target sample may be illuminated by a first plurality of nominal wavelengths
  • a second target sample may be illuminated by a second plurality of nominal wavelengths
  • the first and second pluralities of nominal wavelengths may be the same or different.
  • the number of target samples may be any suitable number greater than or equal to two, for example five or more, ten or more, fifty or more, or one hundred or more.
  • the number of nominal wavelengths by which each sample is illuminated may be any suitable number greater than or equal to two, for example, five or more, ten or more, or twenty or more.
  • the target distribution Do may include a plurality of absorbance spectra respectively associated with the plurality of target samples.
  • Each of the absorbance spectra of the target distribution Do may include the plurality of absorbance values of the associated target sample as measured by the target sensor and respectively associated with the plurality of nominal wavelengths.
  • the second process 200 may include accessing data of a plurality of comparison distributions Di ... DN.
  • This data may be stored in the cloud or in a memory external to the target sensor, and this data may be accessed, for example, by a processing circuit.
  • the processing circuit may be in the target sensor or in a computer or device external to the target sensor, and the processing circuit may be configured to perform the second process.
  • Each of the comparison distributions Di ... DN may be associated with a comparison sensor and may include a plurality of absorbance spectra.
  • each absorbance spectrum of the comparison distribution may include a plurality of absorbance values of a respective sample within the same class of samples as the target samples, measured by the comparison sensor associated with the comparison distribution, and respectively associated with a plurality of reported wavelengths.
  • the plurality of comparison sensors may be sensors that were fabricated similarly or identically to the target sensor. Therefore, even if some of the comparison sensors are miscalibrated, it may be assumed that most of the comparison sensors are accurately or nearly accurately calibrated, or at least that the calibration error varies randomly, from comparison sensor to comparison sensor, and that the mean calibration error, at each wavelength, over all of the comparison sensors, is close to zero.
  • the absorbance data of the samples of the comparison distributions may represent the true absorbance spectra of the target samples when considered collectively, because the samples of the comparison distributions are in the same class of samples as the target samples. Accordingly, the difference between the target distribution Do and the collective comparison distributions Di ... DN characterizes the extent by which the target sensor is miscalibrated.
  • the second process 200 may therefore include comparing the target distribution Do and the comparison distributions Di ... DN to determine how different the target distribution Do is from the comparison distributions Di ... DN.
  • a metric M is utilized to define or measure a distance the target distribution Do and the comparison distributions Di ... DN (i.e., M(Do
  • the metric may be used to define or measure a distance between the target distribution Do and a secondary comparison distribution DI..N (i.e., M(Do
  • a “secondary comparison distribution” is a distribution that is calculated or defined based on a plurality of distributions and may be referred herein as a “comparison distribution of spectra.”
  • the secondary comparison distribution DI..N may be a distribution that is calculated or defined based on the comparison distributions Di ... DN.
  • the secondary comparison distribution DI..N is defined as the union of the comparison distributions Di ... DN, e.g., it is a distribution including the data points from each of the comparison distributions Di ... DN.
  • the metric M is the Wasserstein distance between the target distribution Do(A) and the secondary distribution DI..N(A).
  • the Wasserstein distance may be calculated, for example, using the Python function scipy. stats. wasserstein_distance.
  • the metric M is the Kullback-Leibler divergence, the Renyi divergence, or the f-divergence.
  • Some metrics, such as the Wasserstein distance may use, for discrete distributions such as Do and DI..N, a measure of distance between points in the distributions (where each point may be a pair of vectors resulting from the measurement of a spectrum, as described above).
  • the distance between two such pairs of vectors may be defined as a measure of the distance between two continuous curves each based on a respective one of the two pairs of vectors.
  • Each continuous curve may be formed from the corresponding pair of vectors by, for example, forming a set of ordered pairs, each ordered pair including an element of the wavelength vector and the corresponding element of the absorbance vector, and fitting a curve (of absorbance as a function of wavelength) to the ordered pairs (for example, using linear interpolation or extrapolation, or, as another example, using a spline such as a cubic spline).
  • the distance between the two continuous curves may then be calculated, for example, as the mean squared error between the curves, or as the root mean squared error between the curves, or as the integral of the absolute value of the difference between the curves.
  • the metric M may be defined to be the distance between a first average spectrum (represented by a pair of vectors) and a second average spectrum (represented by a pair of vectors), the distances between pairs of vectors being defined as described above.
  • the first average spectrum may be an average of all of the spectra in the target distribution
  • the second average spectrum may be an average of all of the spectra in the comparison distributions.
  • a calibration function C*(A) may be determined that maps wavelengths to calibrated wavelengths.
  • the calibration function C*(A) may be determined such that, when the nominal wavelengths of the target distribution Do are calibrated by the calibration function C*(A), the distance between the target distribution Do and the comparison distributions Di ... DN, as measured by the metric M, is reduced or minimized. Because the distance is reduced when the nominal wavelengths of the target distribution Do are mapped to the calibrated wavelengths, it may be assumed that the calibration of the target sensor is improved when nominal wavelengths of the target sensor are mapped to calibrated wavelengths using the calibration function C*(A).
  • the third process 300 may be performed by the processing circuit.
  • the calibration function C*(A) may be determined by selecting a function from a family of preliminary calibration functions C(A), for example, by using an algorithm.
  • the family of functions may be, for example, a family of polynomials.
  • several preliminary calibration functions from the family of preliminary calibration functions C(A) may be analyzed to determine how they affect the distance, as measured by the metric M, between the target distribution Do and the comparison distributions Di ... DN when the nominal wavelengths of the target distribution Do are calibrated by the preliminary calibration functions, and the calibration function C*(A) may be defined to be one of the analyzed preliminary calibration functions, such as the preliminary calibration function that reduces the distance, as measured by the metric M, by the greatest amount from among the analyzed preliminary calibration functions.
  • the calibration function C*(A) may be determined by solving the optimization problem ⁇ DI..DN ⁇ ).
  • the family of preliminary calibration functions C(A) may be defined as a function parametrized by several parameters, and selecting the calibration function C*(A) may include using gradient descent or another numerical minimum-finding algorithm. Such an algorithm may iteratively test different preliminary calibration functions and arrive at a calibration function, C'(A) that approximates C*(A).
  • the iteration may terminate when (i) M(Do(C'(A))
  • a threshold e.g., 0.2, 0.1 , or 0.01
  • a trainable neural network may be used as a functional form for the family of preliminary calibration functions C(A), where the neural network includes one input (a wavelength), one output (a calibrated wavelength), and multiple hidden layers with more than one neuron.
  • the neural network may then be fed information, such as information including or based on the comparison distributions Di ... DN, and trained to output wavelengths such that, when the nominal wavelengths of the target distribution Do are calibrated by the neural network, the distance between the target distribution Do and the comparison distributions Di ... DN, as measured by the metric M, is reduced.
  • FIG. 2A depicts a range of true absorbance spectra of a plurality of target samples according to some embodiments.
  • FIG. 2B depicts a target distribution and a secondary comparison distribution according to some embodiments.
  • FIG. 3A depicts a target absorbance function and a comparison absorbance function according to some embodiments.
  • FIG. 3B depicts a calibrated target absorbance function and the comparison absorbance function according to some embodiments.
  • the plurality of target samples may have true absorbance spectra that are closely grouped together because the target samples are within the same class of samples.
  • FIG. 2A shows a first true absorbance spectrum ATrue(A) of a first target sample, and the remaining true absorbance spectra associated with the remaining target samples may be grouped around the first true absorbance spectrum ATrue(A) and generally within a range demarcated by a maximum absorbance boundary AMax(A) and a minimum absorbance boundary AMin(A).
  • Each of the target samples is illuminated with a first nominal wavelength Ai and a second nominal wavelength A2, and first and second absorbance values respectively associated with the first and second nominal wavelengths A1 and A2 are measured for each of the target samples by the target sensor.
  • FIGs. 3A and 3B illustrate these concepts for a simple case in which the target distribution consists of a single spectrum AT(A) as measured by the target sensor, and the secondary comparison distribution consists of a single spectrum Ac(A) as measured by one of the comparison sensors, and the target sensor and the comparison sensor both measure at the same wavelength for the first nominal wavelength A1.
  • the target distribution consists of a single spectrum AT(A) as measured by the target sensor
  • the secondary comparison distribution consists of a single spectrum Ac(A) as measured by one of the comparison sensors
  • the target sensor is miscalibrated relative to the comparison sensor; the target sensor measures at fa - e, whereas the comparison sensor measures at fa.
  • this results in a difference between two continuous functions (straight lines) fit to the respective discrete spectra.
  • the target sensor may report, for the second wavelength (instead of fa), Ac (which is closer to fa - e, than is fa), and, as a result, the two continuous functions fit to the respective discrete spectra differ less than in the uncalibrated case.
  • the target distribution Do may be represented in a 2- dimensional space having as the first dimension a first absorbance A(Ai) associated with the first nominal wavelength Ai and as the second dimension a second absorbance A(A2) associated with the second nominal wavelength fa.
  • the target distribution Do may have a plurality of absorbance spectra respectively associated with the plurality of target samples, and each absorbance spectrum of the target distribution Do may be a point in the 2- dimensional space having a first coordinate value equal to the first absorbance value of the corresponding target sample and a second coordinate value equal to the second absorbance value of the corresponding target sample.
  • the target samples may each be illuminated with light at K different wavelengths, where K is a positive integer greater than 2, the target distribution may be represented in a K-dimensional space, and each of the absorbance spectrum of the target distribution may be a point in the K-dimensional space.
  • FIG. 2B depicts a secondary comparison distribution DI..N in the 2-dimensional space that is based on the comparison distributions Di ... DN.
  • the secondary comparison distribution DI..N may be the union of the comparison distributions Di ... DN.
  • Each of the comparison distributions Di ... DN is associated with a comparison sensor and includes one or more absorbance spectra.
  • each of the absorbance spectra of the comparison distribution includes a pair of first and second absorbance values of a sample within the same class as the target samples, measured by the comparison sensor, and respectively associated with first and second nominal wavelengths Ai and fa.
  • the target distribution Do may be generally aligned with the secondary comparison distribution DI..N along the first dimension A(Ai). However, because the second nominal wavelength fa of the target sensor is miscalibrated and associated with an actual second wavelength fa - e, the target distributions Do may be generally misaligned from the secondary comparison distribution DI..N along the second dimension A(A2).
  • a metric M may be used to measure or define a distance between the target distribution Do and the secondary comparison distribution DI..N.
  • a preliminary calibration function CP(A) may then be selected, for example, from a family of functions C(A) and by an algorithm, to analyze how the distance, as measured by the metric M, between the target distribution Do and the secondary comparison distribution DI..N, changes when the nominal wavelengths of the target distribution Do are calibrated by the preliminary calibration function CP(A).
  • the preliminary calibration function CP(A) may map the first nominal wavelength Ai to itself (i.e. , to Ai) and map the second nominal wavelength fa to a calibrated wavelength Ac.
  • the target distribution Do may then be calibrated by the preliminary calibration function CP(A) to define a calibrated target distribution De.
  • the calibrated target distribution De may include a plurality of calibrated absorbance spectra respectively associated with the plurality of target samples.
  • Each calibrated absorbance spectrum may include the first and second absorbance values of the associated target sample, as measured by the target sensor, but the first and second calibrated absorbance values may be respectively associated with the first and calibrated wavelengths Ai and Ac.
  • a target sensor may be calibrated multiple times over a period time as the target sensor ages and experiences internal changes that may cause the target sensor to become miscalibrated again after an earlier calibration.
  • any instrument configured to measure a quantity that varies as a function of an independent variable may be calibrated, for errors in the setting of the independent variable, utilizing methods and apparatuses within the scope of the present disclosure.
  • a mass spectrometer, a radio frequency spectrum analyzer, an instrument for measuring one or more properties of samples as a function of temperature may be calibrated using analogous methods..

Landscapes

  • Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Mathematical Physics (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)

Abstract

Un procédé de détermination d'une fonction d'étalonnage consiste : à calculer une première distance, entre une distribution de spectre cible et une distribution de spectre de comparaison; à étalonner la distribution de spectre cible à l'aide d'une première fonction d'étalonnage préliminaire afin de former une première distribution de spectre cible étalonnée; à calculer une seconde distance, entre la première distribution de spectre cible étalonnée et la distribution de spectre de comparaison; à déterminer que la seconde distance est inférieure à la première distance; et à régler la fonction d'étalonnage de façon à être égale à la première fonction d'étalonnage préliminaire.
PCT/EP2022/077718 2021-10-06 2022-10-05 Procédés et appareils d'étalonnage de capteur Ceased WO2023057523A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202280067395.XA CN118056115A (zh) 2021-10-06 2022-10-05 用于校准传感器的方法和设备
EP22798136.2A EP4413338A1 (fr) 2021-10-06 2022-10-05 Procédés et appareils d'étalonnage de capteur

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
US202163253054P 2021-10-06 2021-10-06
US63/253,054 2021-10-06
US17/886,409 US12066327B2 (en) 2021-10-06 2022-08-11 Methods and apparatuses for calibrating a sensor
US17/886,409 2022-08-11

Publications (1)

Publication Number Publication Date
WO2023057523A1 true WO2023057523A1 (fr) 2023-04-13

Family

ID=84045094

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2022/077718 Ceased WO2023057523A1 (fr) 2021-10-06 2022-10-05 Procédés et appareils d'étalonnage de capteur

Country Status (2)

Country Link
EP (1) EP4413338A1 (fr)
WO (1) WO2023057523A1 (fr)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0692703A1 (fr) * 1994-07-15 1996-01-17 Unicam Limited Etallonage en longueurs d'ondes d'un spectromètre optique
US20030078746A1 (en) * 1999-11-23 2003-04-24 James Samsoondar Method for calibrating spectrophotometric apparatus
US20080297796A1 (en) * 2007-05-30 2008-12-04 Roche Diagnostics Operations, Inc. Method for the wavelength calibration of a spectrometer
WO2013163268A1 (fr) * 2012-04-24 2013-10-31 Westco Scientific Instruments, Inc. Étalonnage de référence secondaire de spectromètre
WO2017076228A1 (fr) * 2015-11-04 2017-05-11 清华大学 Procédé assisté par ordinateur d'étalonnage en longueur d'onde de spectromètre à bande pleine onde

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0692703A1 (fr) * 1994-07-15 1996-01-17 Unicam Limited Etallonage en longueurs d'ondes d'un spectromètre optique
US20030078746A1 (en) * 1999-11-23 2003-04-24 James Samsoondar Method for calibrating spectrophotometric apparatus
US20080297796A1 (en) * 2007-05-30 2008-12-04 Roche Diagnostics Operations, Inc. Method for the wavelength calibration of a spectrometer
WO2013163268A1 (fr) * 2012-04-24 2013-10-31 Westco Scientific Instruments, Inc. Étalonnage de référence secondaire de spectromètre
WO2017076228A1 (fr) * 2015-11-04 2017-05-11 清华大学 Procédé assisté par ordinateur d'étalonnage en longueur d'onde de spectromètre à bande pleine onde

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
ZHANG LIANG ET AL: "Wavelength calibration based on back propagation neural network", 2014 INTERNATIONAL CONFERENCE ON ANTI-COUNTERFEITING, SECURITY AND IDENTIFICATION (ASID), IEEE, 12 December 2014 (2014-12-12), pages 1 - 4, XP032749658, ISBN: 978-1-4799-7117-6, [retrieved on 20150320], DOI: 10.1109/ICASID.2014.7064972 *

Also Published As

Publication number Publication date
EP4413338A1 (fr) 2024-08-14

Similar Documents

Publication Publication Date Title
US10648840B2 (en) Spectrometer secondary reference calibration
AU2007208473A1 (en) Measuring apparatuses and methods of using them
CN104990501B (zh) 一种三维激光扫描装置的系统参数校准方法
UA86820C2 (uk) Спосіб створення незалежних багатомірних градуювальних моделей
CN114441469A (zh) 水分仪的校准方法、装置和计算机设备
US12066327B2 (en) Methods and apparatuses for calibrating a sensor
CN106680238A (zh) 基于红外光谱分析物质成分含量的方法
CN104502306B (zh) 基于变量重要性的近红外光谱波长选择方法
CN110132416B (zh) 宽波段遥感器在轨光谱定标方法及装置
WO2023057523A1 (fr) Procédés et appareils d'étalonnage de capteur
CN113795748A (zh) 用于配置光谱测定装置的方法
CN118056115A (zh) 用于校准传感器的方法和设备
Watté et al. Computational optimization of the configuration of a spatially resolved spectroscopy sensor for milk analysis
US20180262279A1 (en) Measuring system and measuring method for calibrating an antenna array
CN117390593B (zh) 一种基于辐射计数据的海表参量确定方法以及设备
CN109655415A (zh) 一种波长偏移校正方法及装置及计算机设备
JP7014701B2 (ja) 光学分析装置、並びに光学分析装置に用いられる機械学習装置及びその方法
Engstrøm et al. Analyzing near-infrared hyperspectral imaging for protein content regression and grain variety classification using bulk references and varying grain-to-background ratios
US11353493B2 (en) Data-driven misregistration parameter configuration and measurement system and method
JP2003511687A (ja) 波長較正の最適化方法
Workman Jr First principles of instrument calibration
CN119555618B (zh) 一种抗干扰的宽带腔增强吸收光谱反演方法
JP7785579B2 (ja) 培養液のイオン濃度測定装置
CN119916282B (zh) 电流测试装置自校准方法装置及设备
US11272593B2 (en) System and methods for providing a uniform and tunable hyper-spectral source

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22798136

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 202280067395.X

Country of ref document: CN

WWE Wipo information: entry into national phase

Ref document number: 2022798136

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2022798136

Country of ref document: EP

Effective date: 20240506