Method of comparing spectra to identify similar materials

• US 4,766,551 A
• Filed: 09/22/1986
• Issued: 08/23/1988
• Est. Priority Date: 09/22/1986
• Status: Expired due to Term

• Current Assignee: Pacific Scientific Company 1350 South State College Boulevard Anaheim
• Original Assignee: Pacific Scientific Company Incorporated (Regal Rexnord Corp.)
• Inventors: Begley, Timothy H.
• Primary Examiner(s): NOT, DEFINED
• Assistant Examiner(s): Mattson, Brian M.
• Application Number:
  US06/909,952
• Time in Patent Office: 701 Days
• Field of Search: 364/498, 364/525, 364/526, 364/507, 356/326, 356/388-390, 356/445, 356/446, 356/448, 250/306-308, 250/334, 250/340-342
• US Class Current: 702/28
• CPC Class Codes: G01J 2003/2833 and memorised spectra collection
• CPC Class Codes: G01J 3/28 Investigating the spectrum using colour filters G01J3/51
• CPC Class Codes: G01N 21/359 using near infrared light
• CPC Class Codes: G01N 21/4738 Diffuse reflection precedence is given to G01N21/55 - G01N21/57 if specular component is taken into consideration, e.g. also for testing fluids, fibrous materials

Abstract

In a method of identifying or comparing compositions of material, the reflectivity of a material to be identified or compared is measured at index points distributed through the near infrared spectrum. A similar measurement is made for at least one standard or known material. From the reflectivity measurements, sets of values representing the first or higher order derivative curves are determined. These values mathematically define vectors by representing the coordinates of the end points of the vectors and multiple dimensional space. An index of a similarity between the composition of a test sample and a standard material is calculated by determining the cosine of the angle between the corresponding vectors.

Claims

1. A method of determining whether the composition of a test material is the same as the composition of a standard material, said test material and said standard material each having a reflectivity, comprising the steps of measuring the reflectivity of the standard material throughout a range of the near infrared spectrum to produce a measurement of reflectivity, determining from the measurement of the reflectivity of the standard material a first set of n values representing a derivative of any variation of reflectivity with wavelength throughout said range, measuring the reflectivity of said test material throughout said range to produce a measurement of reflectivity, determining from the measurement of the reflectivity of said test material a second set of n values representing a derivative of any variation of the reflectivity of said test material with wavelength throughout said range, defining a first vector having an end point in n-dimensional space represented mathematically by said first set of n values as coordinates of the end point of said first vector, defining a second vector having an end point in n-dimensional space represented mathematically by said second set of n values as the coordinates of the end point of said second vector, and determining the cosine of any angle between first and second vectors as an index of similarity of the composition of said test material and the composition of said standard material.

2. A method as recited in claim 1, wherein said first and second vectors have magnitudes, and wherein said cosine is determined by determining the vector dot product of said first vector and said second vector divided by the product resulting from multiplying the magnitudes of said first and second vectors.

3. A method as recited in claim 1, wherein said first set of values consists of amplitudes of the derivative determined from the measurement of the reflectivity of said standard material at spaced index points throughout said range and wherein said second set of values consists of amplitudes of the derivative determined from the measurement of the reflectivity of said test material at said spaced index points throughout said range.

4. A method as recited in claim 1, wherein said first set of values are coefficients of terms of a Fourier series representing the derivative determined from the measurement of the reflectivity of said standard material and wherein said second set of values are coefficients of terms of a Fourier series representing the derivative determined from the reflectivity measurement of said test material.

5. A method as recited in claim 1, wherein said steps of measuring reflectivity of the test material, determining a second set of values, defining a second vector, and determining the cosine are performed on each of a plurality of batches of test material to determine a plurality of cosine values, one for each of said batches and segregating the batches for which the cosine value determined is less than a predetermined minimum value from the batches of material for which the cosine value determined is equal to or greater than said predetermined minimum value.

6. A method as recited in claim 5, wherein said predetermined minimum value is determined by measuring the reflectivity of said standard material in said range a plurality of times, determining a plurality of sets of n values each representing a derivative of the variation of reflectivity of said standard material throughout said range measured by a different one of said plurality of measurements, said first set of values being average values determined from said plurality of measurements, defining a vector in n-dimensional space represented mathematically by each set of values of said plurality of sets of n values as the coordinates of the end point of such vector whereby a plurality of vectors are defined corresponding to said plurality of measurements, determining the cosine between each vector of said plurality of vectors and said first vector to provide a plurality of cosine values and selecting the smallest value of said plurality of cosine values as said predetermined minimum value.

7. A method of identifying the composition of an unknown material comprising storing a multiplicity of sets of values each determined from and corresponding to a different one of a multiplicity of materials of different compositions, one of which is the same composition as the composition of said unknown material, each set of values representing a derivative of any variation in reflectivity with wavelength of the corresponding one of said multiplicity of materials throughout a range in the near infrared spectrum, measuring the reflectivity of said unknown material througout said range, determining a further set of values from the measurement of the reflectivity of said unknown material, said further set of values representing a derivative of any variation in the reflectivity of said unknown materiall with wavelength throughout said range, defining a vector having an end point for each set of values of said multiplicity and represented mathematically by such set of values as coordinates of the end point of such vector in multidimensional space, thereby defining a multiplicity of vectors, one for each of said multiplicity of materials, defining a further vector having an end point and represented mathematically by the values of said further set of values as coordinates of the end point of said further vector in multidimensional space, determining the cosine of any angle between said further vector and each vector of said multiplicity of vectors to thereby determine a multiplicity of cosine values, one for each of said multiplicity of materials and identifying the composition of said unknown material as the same as that for which a highest cosine value is determined in said step of determining the cosine of the angle between said further vector and each vector of said multiplicity of vectors.

8. A method as recited in claim 7, wherein said vectors have magnitudes and said cosines are determined by determining the vector dot product of said further vector and each vector of said multiplicity of vectors divided by the product resulting from multiplying the magnitudes of said further vector and such vector of said multiplicity of vectors.

9. A method as recited in claim 7, wherein said multiplicity of sets of values consist of amplitudes of the derivatives corresponding to said multiplicity of materials sampled at spaced index points throughout said range and wherein said further set of values consists of amplitudes of the derivatives determined from the measurement of the reflectivity of said unknown material at said spaced index points throughout said range.

10. A method as recited in claim 7, wherein said multiplicity of sets of values consist of coefficients of terms of a Fourier series representing the derivatives corresponding to said multiplicity of materials and wherein said further sets of values consist of coefficients of terms of a Fourier series representing the derivative determined from the reflectivity measurement of said unknown material.