System Based on Computational Intelligence for Ophthalmology Image Understanding

In this project a computational system of images analysis was developed based on machine learning techniques to aid the diagnosis in the optometry `area`, particularly, an objective and automatic system of ocular refraction errors measurement (astigmatism, hypermetropia and short-sightedness). The results of the work suggest a way to improve the images interpretation from the acquisition technique called Hartmann-Shack (HS) to allow that, later, other ocular problems are detected and measured. The work was realized in an image understanding `area` using Support Vector Machines (SVM). The motivation to investigate images learning techniques for the recognition and analysis of the images in this project was the search for a measurement system capable to interpret the content of the images as a whole, instead of measuring for the comparison of extracted discreet data of the image with extracted data of a reference image.