How to identify this kind of CAPTCHA in a crawler?

Mar.04,2021

No matter what library or machine learning is used, the recognition effect is not good for the current CAPTCHA, so don't put too much effort into it. If there is a real need, look for manual


to extract pixels of the specified color, and then identify them


.

simple text recognition uses OCR libraries, like this https://github.com/zzvirgil/OCR

.

too complex recognition requires too much preprocessing, and most of them don't work well


for this anti-crawler, it is recommended to use account pool, that is, cookie that saves multiple accounts after verification.
there are two other ways to grab packets on PC or APP with wireshark.

Menu