This project boils down to teaching an artificial intelligence which images are what and then we will create "blurred" images so that the artificial intelligence, after being trained, will be able to identify what is
In the identification algorithm, a methodology is used to identify, just like in captchas, what are boats, trucks, cars, etc... and saving this information that is "learned" by the AI in a .keras file
This is where we request the information learned by the AI which is given a % of correct and incorrect results, after that with images that have been transformed into 32 pixels, the AI will analyze byte by byte and then give an answer as to which conclusion it reached