Recognition of merged characters in text based CAPTCHAs

CAPTCHA (Completely Automated Public Turing test to tell Computers and Human Apart) is a type of test to differentiate between humans and computer programs. CAPTCHA attempts to provide security against bots. Although many CAPTCHAs including text, image, audio, and video have been developed in last decade or so, but still these CAPTCHAs are vulnerable against attacks by malicious users. Due to better usability and easy implementations, text based CAPTCHAs are still more popular and powerful mechanism against these attacks. Numerous techniques have been developed to secure text based CAPTCHAs including background noise, random lines, text distortion and merging of characters etc, but still these are vulnerable against bots. In this paper we first discuss these resistance techniques and attacks against these techniques and finally we have proposed an efficient algorithm to recognize the merged character CAPTCHAs.