System and device for preliminary diagnosis of ocular diseases

1. A system for the preliminary diagnosis of ocular diseases, the system comprising:

a camera;

a flash;

a screen for displaying at least one image;

a memory for storing data and a computational application; and

a processor coupled to the camera, the flash, the screen and the memory, wherein upon execution of the computational application by the processor, the processor:

controls the camera and the flash to capture a plurality of images of the eyes of an individual, wherein the plurality of images have a same orientation; and

processes said plurality of images to obtain a final corrected image by:

estimating an amount of light contained in each image of the plurality of images;

discarding one or more respective images that contain no flash light, based on the estimated amount of light contained in each image of the plurality of images, thereby obtaining a plurality of remaining images;

detecting, in a first image of the plurality of remaining images, an approximate area of the face of the individual using an appropriate haar cascade;

cropping said approximate area from said plurality of remaining images, thereby obtaining a plurality of cropped images;

performing a camera stabilization process on said plurality of cropped images, thereby obtaining a plurality of stabilized cropped images;

obtaining an average position of the center of the pupil of each eye of the individual and an average area enclosing each eye of the individual from said plurality of stabilized cropped images; and

selecting one stabilized cropped image of the plurality of stabilized cropped images that has the greatest sharpness as the final corrected image.

2. The system of claim 1 , wherein:

the memory further stores images of previously diagnosed clinical cases with ocular diseases; and

upon execution of the computational application by the processor, the processor:

calculates, in the final corrected image, percentages of respective colors that compose the pupillary reflex of each eye; and

compares the calculated percentages of the respective colors with values obtained from the images of previously diagnosed clinical cases.

3. The system of claim 1 , wherein the system comprises a smart phone that includes the camera.

4. The system of claim 3 , wherein the smart phone further includes the flash.

5. A method for the preliminary diagnosis of ocular diseases, the method comprising:

capturing a plurality of images of the eyes of an individual with a camera using a flash, wherein the plurality of images have a same orientation;

processing the plurality of images, to obtain a final corrected image of the eyes of the individual; and

displaying said final corrected image on a display screen;

wherein processing the plurality of images comprises:

estimating an amount of light contained in each image of the plurality of images;

discarding one or more respective images that contain no flash light, based on the estimated amount of light contained in each image of the plurality of images, thereby obtaining a plurality of remaining images;

detecting, in a first image of the plurality of remaining images, an approximate area of the face of the individual using an appropriate haar cascade;

cropping said approximate area from said plurality of remaining images, thereby obtaining a plurality of cropped images;

performing a camera stabilization process on said plurality of cropped images thereby obtaining a plurality of stabilized cropped images;

obtaining an average position of the center of the pupil of each eve of the individual and an average area enclosing each eye of the individual from said plurality of stabilized cropped images; and

selecting one stabilized cropped image of the plurality of stabilized cropped images that has the greatest sharpness as the final corrected image.

6. The method of claim 5 , wherein:

performing the camera stabilization process on said plurality of cropped images comprises aligning the plurality of cropped images by edge detection and spatial translation to obtain the plurality of stabilized cropped images; and

obtaining an average position of the center of the pupil of each eye of the individual and an average area enclosing each eye of the individual from said plurality of stabilized cropped images comprises:

determining respective areas enclosing the eyes of the individual in each image of said plurality of stabilized cropped images; and

obtaining respective locations of the centers of the eyes from each image of said plurality of stabilized cropped images.

7. The method of claim 6 , wherein aligning the plurality of cropped images by edge detection and spatial translation to obtain the plurality of stabilized cropped images comprises:

detecting first edges in a first image of the plurality of cropped images;

detecting the first edges in at least one subsequent image of the plurality of cropped images; and

calculating a translation of the at least one subsequent image based on the detected first edges in the at least one subsequent image.

8. The method of claim 5 , wherein:

estimating an amount of light contained in each image of the plurality of images comprises calculating a luminance of pixels in each image; and

discarding one or more respective images that contain no flash light comprises discarding the one or more respective images based on the calculated luminance.

9. The method of claim 5 , wherein a number of said plurality of remaining images ranges from 10 to 60 images.

10. The method of claim 5 , wherein obtaining said average position of the center of the pupil of each eye of the individual comprises:

calculating the location, in each image of the plurality of stabilized cropped images, of the center of the pupil of each eye to obtain a plurality of positions of the center of the pupil of each eye;

removing outliers from the plurality of positions to obtain a plurality of remaining positions; and

averaging the plurality of remaining positions to get the average position of the center of the pupil of each eye.

11. The method of claim 5 , wherein selecting one stabilized cropped image of the plurality of stabilized cropped images that has the greatest sharpness as the final corrected image comprises:

defocusing each image of the plurality of stabilized cropped images using a Gaussian filter;

calculating a Fast Fourier Transform (FFT) for each defocused image to obtain an image sharpness value that estimates an image sharpness of each image; and

selecting the final corrected image based at least in part on respective image sharpness values.

12. The method of claim 5 , further comprising:

calculating, from the final corrected image, percentages of respective colors in the pupillary reflex of each eye, wherein the respective colors include red, white, orange and yellow colors.

13. The method of claim 12 , further comprising:

determining whether or not the red color is in a range greater than 50% of pixels that compose an area of the pupil of each eye in the final processed image.

14. The method of claim 13 , wherein the red color is in a range greater than 50% of the pixels that compose the area of the pupil, and the method further comprises:

determining whether or not the yellow color and/or the orange color correspond to percentages in a range greater than 10% of the pixels that compose the area of the pupil of each eye in the final processed image.

15. The method of claim 13 , wherein the red color is in a range less than 50% of the pixels that compose the area of the iris and the pupil, and the method further comprises:

determining whether or not the white color corresponds to a percentage greater than 40% of the pixels that compose the area of the pupil of each eye in the final processed image.

16. The method of claim 5 , further comprising:

comparing the final processed image with a plurality of previous images corresponding to previously diagnosed clinical cases,

wherein the previously diagnosed clinical cases represent at least one of normal cases, clinical cases of refractive defects, and cases of other ocular diseases; and

wherein said refractive defects are selected from the group consisting of hyperopia, astigmatism and myopia.

17. The method of claim 16 , wherein said other ocular diseases are selected from the group consisting of organic diseases and functional diseases.

18. The method of claim 5 , further comprising:

processing said final corrected image to obtain a final corrected image with greater focus;

cutting the eyes of the individual from the final corrected image with greater focus, from the average position of the center of the pupil of each eye and said average area enclosing each eye of the individual; and

post-processing the final corrected image with greater focus, to determine percentages of respective colors that compose the pupillary reflex of each eye in the final corrected image with greater focus.

19. A method for the preliminary diagnosis of ocular diseases, the method comprising:

capturing a plurality of images of the eyes of an individual with a camera using a flash;

processing the plurality of images to obtain a final corrected image of the eyes of the individual; and

displaying said final corrected image on a display screen,

wherein processing the plurality of images comprises:

estimating an amount of light contained in each image of the plurality of images;

discarding one or more respective images of the plurality of images that contain no flash light, based on the estimated amount of light contained in each image of the plurality of images, thereby obtaining a plurality of remaining images;

detecting, in a first image of the plurality of remaining images, an approximate area of the face of the individual using an appropriate haar cascade;

cropping said approximate area from said plurality of remaining images, thereby obtaining a plurality of cropped images;

performing a camera stabilization process on said plurality of cropped images, thereby obtaining a plurality of stabilized cropped images;

obtaining an average position of the center of the pupil of each eye of the individual and an average area enclosing each eye of the individual from said plurality of stabilized cropped images; and

selecting one stabilized cropped image of the plurality of stabilized cropped images that has the greatest sharpness as the final corrected image,

wherein:

estimating an amount of light contained in each image of the plurality of images comprises calculating a luminance of pixels in each image;

discarding one or more respective images that contain no flash light comprises discarding the one or more respective images based on the calculated luminance;

a number of said plurality of remaining images ranges from 10 to 60 images; and

the method further comprises calculating, from the final corrected image, percentages of respective colors in the pupillary reflex of each eye, wherein the respective colors include red, white, orange and yellow colors.