Back to results
US 2019/0294858 A1Published application
Pupil Edge Detection in Digital Imaging
AssigneeWelch Allyn, Inc.
InventorsWilliam Niall Creedon, Eric Joseph Laurin, Richard Allen Mowrey
Life of this patent
Estimating the term…
Abstract
A digital imaging system processes digital images of a subject's fundus and/or pupils to determine a pupil edge. Two regions of a digital image are evaluated to determine a threshold value. Typically, the two regions are selected such that each region would usually not include artifacts. The threshold value can then be used to identify a pupil-iris threshold in the digital image. Based on the pupil-iris threshold, pupil edges are identified.
Claims20
A method of identifying a pupil edge in a digital image, the method comprising:
- receiving a digital pupil image;
- generating a mean first portion pixel intensity of a first portion of the digital pupil image;
- generating a mean second portion pixel intensity of a second portion of the digital pupil image;
- averaging the mean first portion pixel intensity and the mean second portion pixel intensity to generate an average pixel intensity;
- determining a modified standard deviation, including:
- calculating a standard deviation of a pixel intensity for each pixel within the first portion and the second portion;
- and
- multiplying the standard deviation by a multiplier;
- generating a threshold value by summing the average pixel intensity and the modified standard deviation;
- and
- using the threshold value, identifying the pupil edge in the digital pupil image.
8 more elements
A method for identifying a pupil in a digital image, the method comprising:
- receiving a digital pupil image;
- identifying a first portion of the digital pupil image;
- identifying a second portion of the digital pupil image;
- determining a pixel intensity for each pixel within the first portion and each pixel within the second portion;
- determining a mean pixel intensity by calculating the mean of the pixel intensity for each pixel within the first portion and each pixel within the second portion;
- determining a modified standard deviation, including:
- calculating a standard deviation of the pixel intensity for each pixel within the first portion and the second portion;
- and
- multiplying the standard deviation by a multiplier;
- generating a threshold value by summing the mean pixel intensity and the modified standard deviation;
- using the threshold value, identifying a pupil edge in the digital pupil image;
- and
- using the pupil edge, identifying the pupil in the digital pupil image.
10 more elements
A medical imaging system, comprising:
- an illumination assembly including a near-infrared lighting unit;
- a digital camera assembly;
- a processing unit;
- and
- memory storing instructions that, when executed by the processing unit, cause the medical imaging system to:
- illuminate the near-infrared lighting unit;
- receive a digital pupil image with the digital camera assembly;
- obtain a pixel grid size having a pixel grid width and a pixel grid length;
- generate a mean first portion pixel intensity, including evaluating a first portion of the digital pupil image, a first portion length being equal to the pixel grid length and a first portion width being equal to the pixel grid width;
- generate a mean second portion pixel intensity, including evaluating a second portion of the digital pupil image, a second portion length being equal to the pixel grid length and a second portion width being equal to the pixel grid width,
- average the mean first portion pixel intensity and the mean second portion pixel intensity, thereby generating an average pixel intensity;
- determine a modified standard deviation, including:
- calculate a standard deviation of a pixel intensity for each pixel within the first portion and the second portion;
- and
- multiply the standard deviation by a multiplier;
- generate a threshold value by summing the average pixel intensity and the modified standard deviation;
- and
- using the threshold value, identify a pupil edge in the digital pupil image.
15 more elements
Dependencies are inferred from the claim text.