Spectroscopic classification of conformance with dietary restrictions

• US 10,444,213 B2
• Filed: 01/09/2017
• Issued: 10/15/2019
• Est. Priority Date: 08/25/2016
• Status: Active Grant

• Current Assignee: Viavi Solutions, Inc. (Lumentum Holdings Inc.)
• Original Assignee: Viavi Solutions, Inc. (Lumentum Holdings Inc.)
• Inventors: O'Brien, Nada A., Hsiung, Changmeng
• Primary Examiner(s): Huynh, Phuong
• Application Number:
  US15/401,669
• Publication Number:
  US 20180059084A1
• Time in Patent Office: 1,009 Days
• CPC Class Codes: G01N 21/27 using photo-electric detection G01N21/31 takes precedence; circuits for computing concentration logarithmic circuits G06G7/24; photometric circuits in general G01J
• CPC Class Codes: G01N 33/12 Meat; fish
• CPC Class Codes: G06F 18/211 Selection of the most significant subset of features
• CPC Class Codes: G06F 18/2411 based on the proximity to a decision surface, e.g. support vector machines
• CPC Class Codes: G06V 30/242 Division of the character sequences into groups prior to recognition; Selection of dictionaries

Abstract

A device may receive a classification model generated based on a set of spectroscopic measurements performed by a first spectrometer. The device may store the classification model in a data structure. The device may receive a spectroscopic measurement of an unknown sample from a second spectrometer. The device may obtain the classification model from the data structure. The device may classify the unknown sample into a Kosher or non-Kosher group or a Halal or non-Halal group based on the spectroscopic measurement and the classification model. The device may provide information identifying the unknown sample based on the classifying of the unknown sample.

Claims

1. A device, comprising:

a memory; and

one or more processors to;

receive information identifying a result of a spectroscopic measurement of an unknown sample;

perform one or more classifications of the unknown sample based on the result of the spectroscopic measurement and one or more classification models,

the one or more classification models relating to a set of groups,

a first subset of the set of groups being a first meta-group including a particular group, and

a second subset of the set of groups being a second meta-group not including the particular group;

determine a first probability that the unknown sample is associated with the particular group and a second probability that the unknown sample is associated with another group, of the first meta-group, based on the one or more classification models;

classify the unknown sample into the first meta-group based on the first probability and the second probability; and

provide information indicating a classification of the unknown sample into the first meta-group based on classifying the unknown sample into the first meta-group.

2. The device of claim 1, where the one or more processors are further to:

obtain the one or more classification models from a data structure or another device; and

where the one or more processors, when performing the one or more classifications, are to;

perform the one or more classifications based on obtaining the one or more classification models from the data structure or the other device.

3. The device of claim 1, where the one or more processors are further to:

classify the unknown sample into the particular group based on a decision value (DecVal) technique.

4. The device of claim 1,

where the one or more classification models use a support vector machine (SVM) classifier technique, and

where the SVM classifier technique is associated with a linear kernel.

5. The device of claim 1, where the first meta-group and the second meta-group are associated with classification of the unknown sample as Kosher or non-Kosher.

6. The device of claim 1, where the first meta-group and the second meta-group are associated with classification of the unknown sample as Halal or non-Halal.

7. The device of claim 1, where the spectroscopic measurement of the unknown sample is a spectrum corresponding to a cortisol level in the unknown sample,

the cortisol level being associated with inclusion in the first meta-group.

8. A non-transitory computer-readable medium storing instructions, the instructions comprising:

one or more instructions that, when executed by one or more processors, cause the one or more processors to;

receive information identifying a spectrum of an unknown sample analyzed by a spectrometer;

perform a first classification of the unknown sample based on the spectrum of the unknown sample and one or more classification models,

the one or more classification models including a plurality of groups, and

the plurality of groups including;

a first meta-group including a particular group, and

a second meta-group;

determine a first probability that the unknown sample is associated with the particular group and a second probability that the unknown sample is associated with another group, of the first meta-group, based on the one or more classification models;

classify the unknown sample into the first meta-group based on the first probability and the second probability; and

provide information identifying flail the first meta-group to which the unknown sample is classified.

9. The non-transitory computer-readable medium of claim 8, where the first meta-group is a Kosher meta-group or a non-Kosher meta-group.

10. The non-transitory computer-readable medium of claim 8, where the first meta-group is a Halal meta-group or a non-Halal meta-group.

11. The non-transitory computer-readable medium of claim 8, where the one or more instructions, when executed by the one or more processors, cause the one or more processors to:

aggregate information relating to the first meta-group to classify the unknown sample into the first meta-group.

12. The non-transitory computer-readable medium of claim 8, where the one or more instructions, when executed by the one or more processors, cause the one or more processors to classify the unknown sample based on at least one of:

a decision value, or

a confidence metric.

13. The non-transitory computer-readable medium of claim 8,

where the one or more classification models use a support vector machine (SVM) classifier technique, and

where the SVM classifier technique is associated with at least one of:

a linear kernel function type of kernel function,

a radial-basis function type of kernel function,

a sigmoid function type of kernel function,

a polynomial function type of kernel function, or

an exponential function type of kernel function.

14. The non-transitory computer-readable medium of claim 8, where the one or more instructions, when executed by the one or more processors, cause the one or more processors to:

provide information identifying the particular group to which the unknown sample is classified.

15. A method, comprising:

receiving, by a device, a spectroscopic measurement of an unknown sample;

obtaining, by the device, a classification model,

the classification model relating to a set of groups, and

the set of groups including;

a first meta-group including a particular group, and

a second meta-group;

determining, by the device, a first probability that the unknown sample is associated with the particular group and a second probability that the unknown sample is associated with another group, of the first meta-group, based on the classification model;

classifying, by the device, the unknown sample into the first meta-group based on the first probability and the second probability; and

providing, by the device, information identifying the unknown sample based on the classifying of the unknown sample.

16. The method of claim 15, where the classification model is generated based on a plurality of known samples,

the plurality of known samples including at least one of a plurality of:

types of meat,

cuts of meat, or

preparations of meat.

17. The method of claim 15, where the classification model is generated using fewer than 100 known samples.

18. The method of claim 15, where the classification model is generated using fewer than 75 known samples.

19. The method of claim 15, where the classification model is generated using fewer than 50 known samples.

20. The method of claim 15,

where the classification model is generated based on a set of spectroscopic measurements performed by a first spectrometer, and

where the spectroscopic measurement of the unknown sample is received from a second spectrometer.