Klasifikace retinálních snímků dětí

Abstract

This bachelor thesis deals with the automatic classification of retinal images of children using deep learning methods. The aim of the thesis is the design and implementation of a system based on neural networks that enables filtering of images with low informative value, recognition of the left and right eye, and binary classification of the biological sex of the child based on a retinal image. In the thesis, models based on convolutional neural networks are proposed and experimentally verified, while their performance and limitations when working with pediatric retinal data are evaluated.