US6095989A — Optical recognition methods for locating eyes

Claim 1

Independent Claim 84 words

A method implemented by a data processor for locating pupils of a subject comprising the steps of: 1) obtaining at least a partial image of a face of a subject containing an image of at least one pupil, 2) cropping a first pixel array having a high probability of containing an iris and pupil of said subject from said image, said pixel array being slightly larger than a naturally occurring iris, 3) verifying that said first pixel array contains an image of said pupil.

Claim 2

Independent Claim 84 words

A method implemented by a data processor for locating pupils of a subject comprising the steps of: 1) obtaining at least a partial image of a face of a subject containing an image of at least one pupil, 2) cropping a first pixel array having a high probability of containing an iris and pupil of said subject from said image, said pixel array being slightly larger than a naturally occurring iris, 3) verifying that said first pixel array contains an image of said pupil.

Claim 3

Dependent Claim 94 words

A method as set forth in claim 1, wherein said step of cropping said first pixel array from said partial image further comprises the steps of: 1) locating a first pixel registering a highest intensity value in said partial image, 2) centering a second pixel array of a selected size about said first pixel, 3) testing an area around said first pixel to determine if said first pixel is part of a corneal spike, 4) if said first pixel is part of a corneal spike, cropping said second pixel array from said partial image.

Claim 4

Dependent Claim 29 words

A method as set forth in claim 2, further comprising the step of repeating the process of claim 4 a selected number of times until said pupil is located.

Claim 5

Dependent Claim 27 words

A method as set forth in claim 2, further comprising the step of smoothing intensity values in said area from which said second pixel array was cropped.

Claim 6

Dependent Claim 139 words

A method as set forth in claim 4, wherein said step of smoothing intensity values further comprises the steps of: 1) for each row of said second array, 2) interpolating intensity values between a pixel at a beginning of said row and a pixel at an end of said row and substituting interpolated values for each pixel position between said pixel at a beginning of said row and said pixel at an end of said row, 3) for each column of said second array, interpolating intensity values between a pixel at a top of said column and a pixel at a bottom of said column and substituting interpolated values for each pixel position between said pixel at said top of said column and said pixel at a bottom of said column, 4) averaging intensity values of intersecting pixel positions.

Claim 7

Dependent Claim 99 words

A method as set forth in claim 1, wherein said step of verifying that said first pixel array contains an image of said pupil further comprises the steps of: 1) developing a mask array containing first features representative of a disk of a similar size as said pupil, 2) comparing said mask array with a comparison array containing second features representative of a disk of a size of a normal pupil, and where a number of pixels in said first features are within a specified range of pixels in said second features, 3) indicating that a pupil is found.

Claim 8

Dependent Claim 27 words

A method as set forth in claim 6, further comprising the step of conditioning said first pixel array to eliminate superfluous artifacts from said first pixel array.

Claim 9

Dependent Claim 48 words

A method as set forth in claim 7, wherein said first features representative of a disk of the pupil are developed by the step of finding a first circumference representative of a circumference of said pupil and comparing said first circumference with a circumference of said normal pupil.

Claim 10

Dependent Claim 44 words

A method as set forth in claim 7, wherein first features representative of a disk of the pupil are developed by the step of finding an area of said pupil and comparing said area of said pupil with an area of said normal pupil.

Claim 11

Dependent Claim 42 words

A method as set forth in claim 7, further comprising the step of including in said mask array third features representative of eyelids, and comparing said third features representative of eyelids with fourth features representative of normal eyelids in said comparison array.

Claim 12

Dependent Claim 27 words

A method as set forth in claim 7, further comprising the step of determining if features representative of a cataract are present in said first pixel array.

Claim 13

Dependent Claim 111 words

A method as set forth in claim 11, further comprising the steps of: 1) converting said first pixel array to an edge array wherein a circumference of said pupil and said features are defined by contiguous edges, 2) assigning a first value to each pixel position within said pupil edges defined by said edge array, 3) assigning a second value to each pixel position within edges of said features, 4) superimposing a disk of closely similar size as said pupil over said pupil, 5) assigning said first value to each pixel position within said disk, and 6) combining said first values and said second values so that said features are distinguished.

Claim 14

Independent Claim 79 words

A method implemented by a data processor for determining condition of binocular alignment of eyes of a subject imaged by a reflex photometer comprising the steps of: 1) locating a corneal reflection, 2) locating a geometric center of said pupillary aperture, 3) determining a relative location of said corneal reflection with respect to said geometric center of said pupillary aperture, with increasing divergence between said location of said specular peak and said geometric center indicating greater probability of strabismus.

Claim 15

Independent Claim 79 words

A method implemented by a data processor for determining condition of binocular alignment of eyes of a subject imaged by a reflex photometer comprising the steps of: 1) locating a corneal reflection, 2) locating a geometric center of said pupillary aperture, 3) determining a relative location of said corneal reflection with respect to said geometric center of said pupillary aperture, with increasing divergence between said location of said specular peak and said geometric center indicating greater probability of strabismus.

Claim 16

Dependent Claim 62 words

A method as set forth in claim 13, wherein said step of determining relative location of said specular peak with respect to said pupillary aperature further comprises the steps of: 1) finding a ratio of divergence between said corneal reflection and said geometric center, 2) comparing said ratio of divergence of said subject to a similar ratio of divergence of normal eyes.

Claim 17

Independent Claim 140 words

A method implemented by a data processor for locating a pupil of at least one eye in an image taken by a reflex photometer and comprising the steps of: 1) selecting a pixel in said image registering a highest intensity level, 2) applying a series of tests to determine if said pixel is part of a corneal reflection, 3) if said pixel is part of said corneal reflection, then cropping a predetermined area sized to contain at least an iris of said eye about said corneal reflection for further analysis, 4) if said pixel is not part of a corneal reflection, then eliminating said pixel and closely adjacent pixels from further consideration, and 5) selecting a next highest intensity pixel for applying said series of tests to determine if said next highest intensity pixel is part of a corneal reflection.

Claim 18

Independent Claim 140 words

A method implemented by a data processor for locating a pupil of at least one eye in an image taken by a reflex photometer and comprising the steps of: 1) selecting a pixel in said image registering a highest intensity level, 2) applying a series of tests to determine if said pixel is part of a corneal reflection, 3) if said pixel is part of said corneal reflection, then cropping a predetermined area sized to contain at least an iris of said eye about said corneal reflection for further analysis, 4) if said pixel is not part of a corneal reflection, then eliminating said pixel and closely adjacent pixels from further consideration, and 5) selecting a next highest intensity pixel for applying said series of tests to determine if said next highest intensity pixel is part of a corneal reflection.

Claim 19

Dependent Claim 43 words

A method as set forth in claim 15, further comprising the steps of: 1) repeating the steps of claim 15 a predetermined number of times, and if the corneal reflection is not located, 2) implementing a different test procedure to locate said eye.

Claim 20

Dependent Claim 57 words

A method as set forth in claim 16, wherein after about three of the highest intensity pixels are considered and the corneal reflection is not found, then reducing an intensity level of higher intensity pixels in said image by a selected amount, affording a greater opportunity of locating said corneal reflection in following loops of said method.

Claim 21

Dependent Claim 79 words

A method as set forth in claim 17, further comprising the step of testing a selected said highest intensity pixel as being a hot pixel by testing adjacent pixels to said selected pixel to see if adjacent said pixels are lower in intensity by a selected function of intensity of said selected pixel, and if so, then designating said selected pixel as a hot pixel and eliminating said hot pixel from further consideration as to containing said corneal reflection.

Claim 22

Dependent Claim 150 words

A method as set forth in claim 17, further comprising the step of testing a selected said highest intensity pixel as being part of an area of higher intensity pixels greater in size than a corneal reflection by steps comprising: 1) centering a small array larger than a normal corneal reflection on said selected highest intensity pixel, 2) calculating an intensity value that is about one-half the intensity value of said selected pixel to develop a half-mamum intensity value, 3) utilizing said half-maximum intensity value as a threshold, counting a number of pixels in said small array greater than said threshold, and where a counted number of pixel intensity values in said small array is greater than a number of corresponding pixels in a normal corneal reflection and less than a selected maximum of corresponding pixels in a normal eye, then accepting said small array as containing a corneal reflection.

Claim 23

Dependent Claim 116 words

A method as set forth in claim 16, wherein if said corneal reflection is located within said predetermined number of times, then implementing steps comprising: 1) cropping an area around said corneal reflection of a size sufficient to contain at least an iris of said eye, 2) retrieving parameters indicative of a normal eye, 3) comparing like said parameters of a cropped said area around said corneal reflection to said parameters of said normal eye, and where a correlation between said parameters of said normal eye and said cropped area around said corneal reflection is greater than about 70%, 4) indicating that a pupil of said eye is located and measuring several parameters of the pupil.

Claim 24

Dependent Claim 70 words

A method as set forth in claim 16, wherein said different test procedure to locate said eye utilizes steps comprising: 1) normalizing said image, 2) developing a normalized impulse response function of a normal eye, 3) developing a cross correlation matched filter response from said normalized image, and 4) selecting a highest intensity peak resulting from said step of cross correlating said matched filter as a location of said eye.

Claim 25

Dependent Claim 64 words

A method as set forth in claim 21, wherein said step of normalizing said image further comprises the steps of: 1) developing an upper threshold wherein intensity levels in said image above said threshold are clipped, 2) developing a lower threshold wherein intensity levels in said image below said threshold are clipped, 3) scaling unclipped said intensity values to values between 0 and 1.

Claim 26

Dependent Claim 60 words

A method as set forth in claim 22, wherein said upper threshold is developed by steps comprising: 1) calculating a mean value of intensity values in said image, 2) calculating a standard deviation of said intensity values in said image, 3) multiplying said standard deviation by a selected multiplication K factor and adding a result of said multiplying to said.

Claim 27

Dependent Claim 131 words

A method as set forth in claim 22, wherein said lower threshold is developed by steps comprising: 1) inverting lower intensity values in said image so that the highest said lower intensity values are set to -1 and lowest intensity values of said lower intensity values are set to 0, 2) adding 1 to all intensity values to shift them back to positive numbers, 3) translating the mean and standard deviation of intensity values of said image into a 0-1 scale, 4) developing said lower threshold by adding the normalized mean to the normalized standard deviation multiplied by a multiplication K factor, 5) clipping any intensity values greater than 1 and scaling intensity values between 0 and 1, 6) again inverting said array and making inverted, scaled intensity values positive values.

Claim 28

Dependent Claim 129 words

A method as set forth in claim 21, wherein said step of developing a normalized response function of a normal eye further includes the steps of: 1) developing a first array having a disk of a radius equal to a normal iris radius, said disk centered in said array and populated with intensity values set to a negative of a constant times a normalized standard deviation, 2) developing a second array having a disk of a radius equal to a normal pupil radius, said disk centered in said array and populated with intensity values double a constant times said normalized standard deviation, 3) adding said first array and said second array together to form a summed array, 4) adding a mean intensity value to develop said normalized response function.

Claim 29

Dependent Claim 122 words

A method as set forth in claim 25, further comprising the steps of: 1) performing a fast Fourier transform on said first array and said second array to develop a first FFT result array and a second FFT result array, respectively, 2) multiplying, element by element, said first FFT result array and said second FFT result array to produce an FFT cross correlation array, 3) applying said FFT cross correlation array to a high pass filter to obtain a filtered array, 4) performing an inverse fast Fourier transform on said filtered array to provide a cross correlation array, and 5) locating a highest intensity location in said cross correlation array, said highest intensity pixel being indicative of a location of said eye.

Annuity Stage	Window Opens	Payment Deadline	Surcharge Period Until	Estimated Fee	Status
3.5 Years (Year 4)	2003-08-01	2004-01-31	2004-08-01	$2000.00	Window Closed / Lapsed
7.5 Years (Year 8)	2007-08-01	2008-01-31	2008-08-01	$3760.00	Window Closed / Lapsed
11.5 Years (Year 12)	2011-08-01	2012-01-31	2012-08-01	$7700.00	Window Closed / Lapsed

Optical recognition methods for locating eyes

Abstract

Inventors

Bibliographic Details

Claims Analysis

References Cited (Backward Citations)

USPTO Annuity & Maintenance Windows

Assigned Categories & Tags

Assign Category Tag