Abstract: A system for determining a refractive error from a red reflex image of the eyes of a subject by analyzing the red reflex image of the subject using deep learning and clustering algorithms is provided. The system (i) predicts a probability map for each of sixteen eye landmarks on a contour of an iris of the eye, (ii) determines a location of an eye landmark on the eye-region image, (iii) determines a circular mask of the iris, (iv) extracts an iris region image between the circular mask of the iris and the eye-region image, (v) generates a plurality of clusters by identifying a similarity between pixels of the iris region image, (vi) determines a crescent of the red reflex in the iris, (viii) predicts a pupil radius, (ix) determines a width of the crescent, and (x) determines a power in each eye by computing a width of an anti-crescent, eccentricity of an image capturing device, a working distance of the image capturing device.

Abstract: A system for determining a spherical power of the eyes of the subject based on the measured refractive error associated with the red reflex image is provided. The system (i) predicts a position of a plurality of facial landmarks on the facial image of the subject, (ii) extracts the eyes from the facial image based on the facial landmarks, (iii) determines a position of the eyelid, (iv) determines a red reflex in the iris of the eye, (v) locates the red reflex with a reddish hue and a crescent in the iris for measuring the refractive error, (vi) generates a mask of the crescent by thresholding the red reflex image, (vii) determines a width of anti-crescent in the red reflex image, and (viii) determines a spherical power using the width of the anti-crescent, eccentricity and working distance of the image capturing device and a diameter of the pupil.

Abstract: Techniques for estimating blepharoptosis from an image of a face are disclosed. The image is captured by a camera. Histogram of Oriented Gradients and Linear Classifiers are used to predict the location of the face and eyes. An Ensemble of Regression Trees is used to detect points on each eyelid and the canthi. The location of the iris is ascertained by thresholding HLS values of individual pixels and applying a sliding window algorithm. Using regression, the eyelid shape curve is approximated. The eyelid shape curve function is used to measure the MRD1 value and distance below the limbus after detecting the Purkinje image/corneal light reflex. Finally, the intensity of blepharoptosis is estimated by categorizing the measured MRD1 and distance-from-Limbus values.

Abstract: Techniques for estimating blepharoptosis from an image of a face are disclosed. The image is captured by a camera. Histogram of Oriented Gradients and Linear Classifiers are used to predict the location of the face and eyes. An Ensemble of Regression Trees is used to detect points on each eyelid and the canthi. The location of the iris is ascertained by thresholding HLS values of individual pixels and applying a sliding window algorithm. Using regression, the eyelid shape curve is approximated. The eyelid shape curve function is used to measure the MRD1 value and distance below the limbus after detecting the Purkinje image/corneal light reflex. Finally, the intensity of blepharoptosis is estimated by categorizing the measured MRD1 and distance-from-Limbus values.