Method and apparatus for digital analysis of multiple component visible fields

Claim 1

A method for digitally analyzing two component fields, comprised of human hair and human scalp, each having a relatively distinct, visible intensity, to measure a relative proportional content of each component comprising: recording a selected field area to create a record having a plurality of discrete segments of varying intensity; digitally resolving the intensity of each of said discrete segments into one of a multiple of discrete values; registering a frequency of occurrence of each said multiple of discrete values in said field to obtain a frequency distribution; and resolving said frequency distribution into multiple, normally distributed component distributions, through a Gaussian mixture analysis, each being representative of a selected component to provide a first measure of a proportional content of said field comprised of each component.

Claim 2

A method for digitally analyzing two component fields, comprised of human hair and human scalp, each having a relatively distinct, visible intensity, to measure a relative proportional content of each component comprising: recording a selected field area to create a record having a plurality of discrete segments of varying intensity; digitally resolving the intensity of each of said discrete segments into one of a multiple of discrete values; registering a frequency of occurrence of each said multiple of discrete values in said field to obtain a frequency distribution; and resolving said frequency distribution into multiple, normally distributed component distributions, through a Gaussian mixture analysis, each being representative of a selected component to provide a first measure of a proportional content of said field comprised of each component.

Claim 6

A method for digitally analyzing multiple-component fields to measure a relative proportional content of each component comprising: recording a selected field area to create a record having a plurality of discrete segments of varying intensity; digitally resolving the intensity of each of said discrete segments into one of a preselected multiple of discrete values; registering a frequency of occurrence of each said preselected multiple of discrete values in said field to obtain a frequency distribution; resolving said frequency distribution into multiple, normally distributed component distributions, through a Gaussian mixture analysis, each being representative of a selected component to provide a first measure of a proportional content of said field comprised of each component; storing a log likelihood for a single Gaussian distribution of said frequency distribution; storing a log likelihood for a two component, mixed Gaussian distribution of said frequency distribution; and comparing said stored values to obtain a difference between said stored values, whereby said field may be approximated by a two-component Gaussian mixture distribution if said stored values differ by a predetermined amount.

Claim 7

A method for digitally analyzing multiple-component fields to measure a relative proportional content of each component comprising: recording a selected field area to create a record having a plurality of discrete segments of varying intensity; digitally resolving the intensity of each of said discrete segments into one of a preselected multiple of discrete values; registering a frequency of occurrence of each said preselected multiple of discrete values in said field to obtain a frequency distribution; resolving said frequency distribution into multiple, normally distributed component distributions, through a Gaussian mixture analysis, each being representative of a selected component to provide a first measure of a proportional content of said field comprised of each component; storing a log likelihood for a single Gaussian distribution of said frequency distribution; storing a log likelihood for a two component, mixed Gaussian distribution of said frequency distribution; and comparing said stored values to obtain a difference between said stored values, whereby said field may be approximated by a two-component Gaussian mixture distribution if said stored values differ by a predetermined amount.

Claim 8

A method for digitally analyzing multiple-component fields to measure a relative proportional content of each component comprising: recording a selected field area to create a record having a plurality of discrete segments of varying intensity; digitally resolving the intensity of each of said discrete segments into one of a preselected multiple of discrete values; registering a frequency of occurrence of each said preselected multiple of discrete values in said field to obtain a frequency distribution; resolving said frequency distribution into multiple, normally distributed component distributions, through a Gaussian mixture analysis, each being representative of a selected component to provide a first measure of a proportional content of said field comprised of each component; storing a log likelihood for a single Gaussian distribution of said frequency distribution; storing a log likelihood for a two component, mixed Gaussian distribution of said frequency distribution; and comparing said stored values to obtain a difference between said stored values, whereby said field may be approximated by a two-component Gaussian mixture distribution if said stored values differ by a predetermined amount. repeating each step after a selected observation period to provide a second measure of the proportional content of said field comprised of each component; and comparing said second proportional content measure with the first proportional content measure to register any change in the proportional content of each component.

Claim 9

A method for digitally analyzing multiple-component fields to measure a relative proportional content of each component comprising: recording a selected field area to create a record having a plurality of discrete segments of varying intensity; digitally resolving the intensity of each of said discrete segments into one of a preselected multiple of discrete values; registering a frequency of occurrence of each said preselected multiple of discrete values in said field to obtain a frequency distribution; resolving said frequency distribution into multiple, normally distributed component distributions, through a Gaussian mixture analysis, each being representative of a selected component to provide a first measure of a proportional content of said field comprised of each component; storing a log likelihood for a single Gaussian distribution of said frequency distribution; storing a log likelihood for a two component, mixed Gaussian distribution of said frequency distribution; and comparing said stored values to obtain a difference between said stored values, whereby said field may be approximated by a two-component Gaussian mixture distribution if said stored values differ by a predetermined amount. repeating each step after a selected observation period to provide a second measure of the proportional content of said field comprised of each component; and comparing said second proportional content measure with the first proportional content measure to register any change in the proportional content of each component.

Claim 10

A method of monitoring a relative proportional content of each component in a multiple-component visible field comprising: recording a selected visible field to create a first record of said selected visible field, said record having a plurality of discrete portions of varying visible intensity; digitally resolving the visible intensity of each of said discrete portions into one of a preselected multiple of discrete values; determining a frequency of occurrence of each discrete value in said visible field to obtain a frequency distribution; resolving said frequency distribution through a Gaussian mixture analysis into a plurality of multiple, normally distributed component distributions, each of said component distributions being representative of a selected component in said visible field, thereby providing a first proportional measure of said visible field comprised of each component; repeating the above steps after a selected interval to create a second record of said preselected visible field to obtain a second set of multiple, normally distributed component distributions to provide a second proportional measure of the field comprised of each component; and comparing said second proportional measure with said first proportional measure to register any change in the relative proportional content of each component in said selected visible field.

Claim 11

A method of monitoring a relative proportional content of each component in a multiple-component visible field comprising: recording a selected visible field to create a first record of said selected visible field, said record having a plurality of discrete portions of varying visible intensity; digitally resolving the visible intensity of each of said discrete portions into one of a preselected multiple of discrete values; determining a frequency of occurrence of each discrete value in said visible field to obtain a frequency distribution; resolving said frequency distribution through a Gaussian mixture analysis into a plurality of multiple, normally distributed component distributions, each of said component distributions being representative of a selected component in said visible field, thereby providing a first proportional measure of said visible field comprised of each component; repeating the above steps after a selected interval to create a second record of said preselected visible field to obtain a second set of multiple, normally distributed component distributions to provide a second proportional measure of the field comprised of each component; and comparing said second proportional measure with said first proportional measure to register any change in the relative proportional content of each component in said selected visible field.

Claim 12

A method for digitally analyzing multiple-component visible fields to determine a relative proportional content of each component comprising: A. recording a first section of a selected visible field area to create a record of said first section, said first section being comprised substantially of a first-component having a first visible intensity; B. digitally resolving the visible intensity of said first-component into a first of a preselected multiple of discrete values; C. repeating the above steps to identify discrete digital values for the visible intensity of each component in said multiple-component visible field; D. recording said visible field area to create a record of said selected visible field area, said record having a plurality of discrete segments of varying visible intensity; E. digitally resolving the visible intensity of each of said discrete segments into one of said preselected multiple of discrete values; F. registering a frequency of occurrence of discrete segment values within a predetermined range of each of said digital component values, whereby a proportional measure of a content of said visible field comprised of each component is provided by a proportion of said plurality of discrete segments having discrete values within said range of each component.

Claim 13

A method for digitally analyzing multiple-component visible fields to determine a relative proportional content of each component comprising: A. recording a first section of a selected visible field area to create a record of said first section, said first section being comprised substantially of a first-component having a first visible intensity; B. digitally resolving the visible intensity of said first-component into a first of a preselected multiple of discrete values; C. repeating the above steps to identify discrete digital values for the visible intensity of each component in said multiple-component visible field; D. recording said visible field area to create a record of said selected visible field area, said record having a plurality of discrete segments of varying visible intensity; E. digitally resolving the visible intensity of each of said discrete segments into one of said preselected multiple of discrete values; F. registering a frequency of occurrence of discrete segment values within a predetermined range of each of said digital component values, whereby a proportional measure of a content of said visible field comprised of each component is provided by a proportion of said plurality of discrete segments having discrete values within said range of each component.

Claim 15

A method for digitally analyzing a human scalp region to measure a proportion of scalp area comprised of hair relative to that area comprised of bare skin comprising the steps of: recording said scalp region to create a record of said scalp region, said record having a plurality of discrete segments of varying visible intensity; digitally resolving the visible intensity of each of said discrete segments into one of a preselected multiple of discrete values; storing a count of each said discrete value in said scalp region to obtain a frequency distribution; resolving said frequency distribution through a Gaussian mixture analysis into two normally distributed component Gaussian distributions, one of said component distributions being representative of the visible intensity of hair and the other component distribution being representative of the visible intensity of bare skin, thereby providing a measure of a relative proportional content of hair in the scalp region.

Claim 16

A method for digitally analyzing a human scalp region to measure a proportion of scalp area comprised of hair relative to that area comprised of bare skin comprising the steps of: recording said scalp region to create a record of said scalp region, said record having a plurality of discrete segments of varying visible intensity; digitally resolving the visible intensity of each of said discrete segments into one of a preselected multiple of discrete values; storing a count of each said discrete value in said scalp region to obtain a frequency distribution; resolving said frequency distribution through a Gaussian mixture analysis into two normally distributed component Gaussian distributions, one of said component distributions being representative of the visible intensity of hair and the other component distribution being representative of the visible intensity of bare skin, thereby providing a measure of a relative proportional content of hair in the scalp region.