The deliverable is a job/script that takes an image (CAPTCHA) as input, and returns the text displayed in the image as a text. The images have a variable character position/orientation and a lot of noise both from dots and lines. The ideal solution should clean all the noise, and perform some kind of OCR to determine the text. Is is also valid to use machine learning to transform the cleared image in text.
Skills & Expertise RequiredMachine Learning
Offer to work on this project closes in 292 days!Submit A Proposal
Share this project with your friends