AUTOMATIC DETECTION OF DEFECTS IN RAILWAY TRACKS USING COMPUTER VISION

Abstract

Artificial Intelligence has proven to be a promising tool in the field of computer vision. This work proposes the development of a solution using Machine Learning techniques to detect surface defects in railway tracks. Based on a dataset obtained through track images, a model will be developed to identify defects, aiming to enhance safety and efficiency in railway maintenance. Preliminary results show that a classification model was created with satisfactory evaluation metrics, highlighting the potential of this application to assist in railway inspections.