This project implements an automated system capable of segmenting and recognizing characters in distorted text-based CAPTCHA images. The complete pipeline includes web scraping, preprocessing, segmentation, data augmentation, and supervised Machine Learning classification.