A Survey on Breaking Technique of Text-Based CAPTCHA

Abstract

The CAPTCHA has become an important issue in multimedia security. Aimed at a commonly used text-based CAPTCHA, this paper outlines some typical methods and summarizes the technological progress in text-based CAPTCHA breaking. First, the paper presents a comprehensive review of recent developments in the text-based CAPTCHA breaking field. Second, a framework of text-based CAPTCHA breaking technique is proposed. And the framework mainly consists of preprocessing, segmentation, combination, recognition, postprocessing, and other modules. Third, the research progress of the technique involved in each module is introduced, and some typical methods of segmentation and recognition are compared and analyzed. Lastly, the paper discusses some problems worth further research.

Benchmarks for Designing a Secure Devanagari CAPTCHA

“…CAPTCHAs are available in many forms like text based, image based, audio based, puzzle based and video based. Most of the text-based CAPTCHAs are broken with high success [10,11]. In 2003, Mori et al successfully attacked Gimpy CAPTCHA with breaking rate 33-92% and EZ-Gimpy CAPTCHA with breaking rate 92% using the shape context matching technique [12].…”

A Low-Cost Approach to Crack Python CAPTCHAs Using AI-Based Chosen-Plaintext Attack

“…The ensemble learning approach was adopted to vote the recognition scores, which is robust to noise [17]. Chen et al classified the CAPTCHA segmentation methods into several categories such as projection, connection, width, contour, structure, and filter, and highlighted the character recognition methods based on neural network and deep learning tactics [7].…”

“…Inspired by the previous segmentation algorithms [4,7,24], we proposed our own peak segmentation approach that appears to be our best bet at generating large amounts of training data while keeping the bounding boxes relatively close to the character text. An example is shown in Figure 4 to visually observe how this peak segmentation process breaks down the image and reduces the bounds further.…”

“…There are many schemes including proprietary CAPTCHA schemes such as Google reCAPTCHA and Yahoo CAPTCHA [7], and open-source CAPTCHA libraries such as Python Claptcha (https: //pypi.org/project/claptcha/) and Captcha ( we differentiate CAPTCHA and Captcha. The uppercase one represents the authentication image while the latter one stands for the open-source library/module, https://pypi.org/project/captcha/).…”

“…Claptcha-like CAPTCHAs are widely adopted for Internet service authentications by many profit or non-profit institutions such as Delta Airline and MITBBS. Since the proprietary CAPTCHA schemes were used by well-known companies such as Google and Yahoo, these types of CAPTCHA schemes have drawn more attentions in the security community and many existing CAPTCHA solvers aim to crack these proprietary CAPTCHA schemes [7,8]. It also results in that the proprietary CAPTCHA schemes add more techniques to increase their security.…”