基于Web搜索引擎的CAPTCHA构造方法实现

摘要全自动开放式人机区分图灵测试（CAPTCHA）俗称验证码技术，在互联网服务中被广泛应用于防止恶意程序访问和滥用服务资源。传统的CAPTCHA构造方法采用文本识别的方法，通过显示进行了一定变异和模糊的图片，要求用户对其中的文字进行识别和输入。随着字符分割和光学识别技术(OCR)的发展，绝大多数文本验证码都可以通过模式识别和自然语言理解的方法加以破解。本项目研究并实现一种基于互联网搜索引擎的全新的验证码构造技术，利用搜索引擎的图片搜索能力，对于随机关键字搜索并选择少量相关图片，混杂在大量非相关图片中。要求用户针对关键字，识别并正确选择出相关图片。由于采用了互联网信息作为背景资料库，采用搜索引擎作为资料选择和判断依据，本方法具有动态的、海量的验证码构造结果，可以抵抗攻击者的暴力破解或者穷举尝试。25375
关键词图灵测试逆测试图片搜索引擎图像验证码随机关键字
毕业论文设计说明书外文摘要
Title   The Design and Implementation of CAPTCHA Construction
           System Based on Web Search Engine
Abstract
Completely automated public Turing test to tell computers and humans apart (CAPTCHA), commonly known as the verification code technology, are widely used to prevent malicious programs from accessing and abusing of service resources in internet service. Traditional CAPTCHA constructions use the method of text recognition by displaying pictures with a certain amount of variation and fuzzy, and require users to identify the characters and then input them. With the development of character segmentation and optical character recognition (OCR), the majority of text CAPTCHA can be cracked by the method of pattern recognition and natural language understanding. The project designs and implements a new CAPTCHA construction based on web search engines, which random keyword searches and select a few related images with a random keyword by the image-searching capability of the search engine. The keyword-related pictures are mixed with a large number of unrelated pictures. It asks users identifying and selecting the related picture. As a result of the background information database from internet information, and the material selection and judgment basis from web search engine, this method has the dynamic mass of results of the code structure, which can resist the attacker's brute force or exhaustive attempt.
Keywords inverse Turing test   web search engine   image CAPTCHA random keyword
目   次
1 引言    1
1.1 课题背景    1
1.2 课题内容    1
1.3 研究意义    1
2 研究现状与相关技术    3
2.1 研究现状    3
2.2 相关技术    6
3 系统设计    11
3.1 算法描述    11
3.2 模块设计    13
4 系统实现与测试    17
4.1 系统实现    17
4.2 性能测试分析    20
结论    26
致谢    27
参考文献28
1 引言
1.1 课题背景
随着互联网的迅速普及和计算机网络技术的飞速发展，人机交互模式成为现代生活中不可或缺的部分，如在线订单、论坛发帖等。然而，互联网在方便了人们的工作和生活的同时，也带来了潜在的安全隐患。常有恶意的自动软件会滥用这些面向人类用户的网络服务，从而可能导致用户的个人信息被泄露、权益被侵犯。基于Web搜索引擎的CAPTCHA构造方法实现:http://www.751com.cn/jisuanji/lunwen_19125.html