Image Recognition and Annotation based Decision Making of CAPTCHAs for Human Interpretation
S. Ezhilarasi, P. Uma Maheswari · 2020
In current scenario of web based security, CAPTCHA security is trending and widely used in many applications. The major functional requirement of CAPTCHA is that it should be easy for the human interpretation but difficult for the bots to identify it. This feature leads to an intellectual effort for gaining accuracy. The proposed system is an image CAPTCHA which provides attractive interaction with the users that they could solve it without assistance of anyone. The main motive of this paper is to demonstrate that the CAPTCHA is secure and cannot be cracked and easier for Image Recognition and text annotation in decision-making of users to select and annotate the images where the bots cannot. The distorted images can be recognizable by users which are harder for bots that differentiates human and bots without affecting the human recognition. In this paper, selection based image CAPTCHA is dealt and an interactive and attractive image recognition and annotation based decision-making model is proposed in which 85% of participants felt easy in solving it and obtained a success rate of 97% in login access and also prevents DDoS attacks.