网上银行验证码研究与安全性分析

验证码（CAPTCHA）,是用于区别用户是人类还是机器的一种计算机自动程序。作为一种辅助手段,验证码在互联网安全领域扮演着很重要的角色。为了对网上银行验证码进行安全性评价,其过程包括三方面:验证码图像采集、图像预处理和图像识别。对国内网上银行验证码特点进行分析,采用最具可靠性的BP神经网络算法,选取其中具有代表性的验证码进行训练、识别。从识别结果中分析,通过所得到的识别率来评价国内网上银行验证码的安全有效性,并对网上银行验证码的生成给出一定的建议。     

基于深度学习的字符型验证码识别研究

字符型验证码作为常见的验证码类型,被广泛应用在各种网络平台,作为一种防止自动化脚本入侵的信息安全手段.针对这种验证码识别问题提出了一种基于卷积神经网络来识别字符型图片验证码的方法.采用TensorFlow深度学习框架对卷积神经网络模型进行训练,将灰度化的验证码图像作为输入,通过验证码数据集进行实验.结果表明,该模型对识别字符型验证码具有较好的泛化能力与鲁棒性.     

基于 OCR 技术的图像验证码识别

验证码是一种标准的网络安全技术,它主要用来防止网民或者黑客等对网站的恶意注册和访问,以及发送垃圾文件、暴力破解高价值密码、滥发广告等恶意事件。通过对图片的扫描,可以提取图片中的数字、字符信息等。本文提出了一种基于OCR技术的图形验证码识别技术。通过对验证码图片进行灰度化、二值化、去噪点、圈点填充、直线填充、图像分割、统一大小、图像匹配和存入字库等一系列的操作,进行对验证码的识别实验。本文通过实验对验证码的特点有了充分的了解,这样方便设计出更加安全的验证码,防止被不法分子破解。     

基于Python和CNN的验证码识别

针对目前互联网上关于页面自动登录环节出现的难点,由于部分登录界面有验证码的存在,自动登录的时长被增加,并且有的验证码难以识别,这就提出了基于Python和卷积神经网络(CNN)相结合的验证码识别。首先本文对三千多张验证码的样本集进行图片预处理,分别有灰度化处理、二值化处理和去噪点处理三步操作。然后利用三个池化层和一个全连接层的结构设计卷积神经网络,随后训练样本集,并对随机的十个样本进行预测。     

基于Tesseract引擎样本训练的验证码识别

笔者利用OCR算法引擎Tesseract的样本训练方法,对简单验证码和复杂验证码进行识别。使用预处理后的单字符图片作为训练样本,对算法进行样本训练,以提高算法对普通字符验证码的识别率,并且使其可以识别较为复杂的字符验证码。实验结果证明,样本训练后的算法对简单验证码的识别率达到了99%以上,且可以对原本几乎无法识别的复杂验证码进行有效识别。     

基于形状上下文的复杂验证码识别算法

使用改进的形状上下文方法对复杂验证码进行识别,采用整体识别方法,不对图片进行切割,使用半极坐标圆进行建模的方式,解决2个字符连接处像素点建模互相干扰的问题。设计并实现复杂验证码识别算法,并与简单验证码进行比较。实验结果证明,复杂验证码识别算法能对字符粘连的复杂验证码进行识别。     

基于对抗网络的验证码识别方法

在人机智能交互中,让机器自动识别验证码是机器模拟人的一项基础技术。基于文本的验证码识别一般先对验证码图片进行预处理,然后切割,最后对字符分类识别。字符切割的准确程度直接影响最终识别结果。提出一种对抗学习方法识别文本型验证码。先训练一个Pix2pix网络对验证码图片进行预处理,然后对抗训练出一对分割和识别网络。分割网络不仅能分割粘贴字符,而且可以筛选出难以分割的验证码结果。识别网络采用上下文相关的多通道卷积网络,能有效解决分割过程中因信息丢失而无法识别的问题。实验结果表明,该方法能提高文本验证码识别的准确率。     

覆盖约简算法在Captcha识别方面的应用

Captcha是用来区分计算机与人类的一种程序,图像验证码是一种典型的Captcha。对图像验证码的发展历程进行了总结,比较了几种目前最常见验证码的特点和设计思路,提出了对一般验证码的破解方法,并设计实现了提取验证码数据的图片信息提取系统;应用基于条件信息熵的覆盖约简算法,对验证码进行识别,对比实验结果证明,识别效果良好。     

基于字符分割和LeNet-5网络的字符验证码识别

为了解决传统验证码识别方法效率低,精度差的问题,设计了一种先分割后识别的验证码处理方案。该方案在预处理阶段用中值滤波去噪,再利用霍夫变换对图像字符进行矫正;在字符分割阶段,利用垂直投影算法确定验证码字符块个数,以及字符坐标点,再用颜色填充算法对验证码进行初步分割,根据分割后的字符块数量对粘连字符进行二次分割;在识别阶段,我们对LeNet-5网络进行了改进,修改了输入层,并用全连接层替换了LeNet-5网络中的C5层,以此来对验证码字符进行识别;实验表明,对于非粘连验证码和粘连验证码,单张图片分割时间为0.14和0.15ms,分割准确率为98.75%和97.25%,识别准确率为99.99%和97.7%;结果表明,该算法对验证码分割和识别都有着很好的效果。     

基于卷积神经网络的验证码识别

验证码识别与设计是目前人工智能领域的挑战性问题,验证码图片内容识别通过强制人机交互来抵御机器自动化攻击的,验证码是否能被批量识别可以用来衡量验证码设计的优劣。目前已经有相对成熟的算法解决这类问题,但是仍然存在天花板有待突破。首先本文对5000张验证码的样本集进行图片预处理,对验证码图片去噪点和切割操作。然后利用添加了注意力模块的卷积神经网络训练样本集,并对另外5000张样本进行预测,测试集的准确率可以达到97.9%。     

CAPTCHA技术研究综述

全自动开放式人机区分图灵测试(completely automated public Turing test to tell computers and humans apart,CAPTCHA)又称为人机交互验证(human interactive proof,HIP),它能自动产生并评估一个测试,这个测试能被几乎所有人类用户通过,而现有的计算机程序不能通过.CAPTCHA提供了一种自动驱分人和机器的手段,已成为一种标准的网络安全技术成功应用于包括Google,Yahoo!以及微软在内的各大网站.CAPTCHA设计基于人工智能领域的开放性问题,按表现载体和内容不同分为文本、图像、声音3种类型,其中,基于字符识别的文本CAPTCHA已得到广泛使用;图像CAPTCHA利用计算机视觉中的难解问题,目前尚处于研究阶段;声音CAPTCHA针对视觉残障者,是对前两种视觉CAPTCHA的补充.介绍了CAPTCHA的发展和设计准则,详细阐述CAPTCHA设计和破解的研究工作及最新进展,给出典型实例,讨论其可用性和安全性,最后指出未来CAPTCHA技术的发展方向和亟待解决的问题.     

Relative image similarity learning with contextual information for Internet cross-media retrieval

With the fast explosive rate of the amount of image data on the Internet, how to efficiently utilize them in the cross-media scenario becomes an urgent problem. Images are usually accompanied with contextual textual information. These two heterogeneous modalities are mutually reinforcing to make the Internet content more informative. In most cases, visual information can be regarded as an enhanced content of the textual document. To make image-to-image similarity being more consistent with document-to-document similarity, this paper proposes a method to learn image similarities according to the relations of the accompanied textual documents. More specifically, instead of using the static quantitative relations, rank-based learning procedure by employing structural SVM is adopted in this paper, and the ranking structure is established by comparing the relative relations of textual information. The learning results are in more accordance with the human’s recognition. The proposed method in this paper can be used not only for the image-to-image retrieval, but also for cross-modality multimedia, where a query expansion framework is proposed to get more satisfactory results. Extensive experimental evaluations on large scale Internet dataset validate the performance of the proposed methods.     

Exploring the influence of CAPTCHA types to the users response time by statistical analysis

CAPTCHA stands for Completely Automated Public Turing Test to Tell Computers and Humans Apart. It is a test program that solves a given task for preventing the attacks made by automatic programs. If the response to CAPTCHA is correct, then the program classifies the user as a human. This paper introduces a new analysis of the impact of different CAPTCHAs to the Internet user’s response time. It overcomes the limitations of the previous approaches in the state-of-the-art. In this sense, different types of CAPTCHAs are presented and described. Furthermore, an experiment is conducted, which is based on two populations of Internet users for text and image-based CAPTCHA types, differentiated by demographic features, such as age, gender, education level and Internet experience. Each user is required to solve the different types of CAPTCHA, and the response time to solve the CAPTCHAs is registered. The obtained results are statistically processed by Mann-Whitney U and Pearson’s correlation coefficient tests. They analyze 7 different hypotheses which evaluate the response time in dependence of gender, age, education level and Internet experience, for the different CAPTCHA types. It represents an invaluable study in the literature to predict the best use of a given CAPTCHA for specific types of Internet users.

     

Segmentation of connected characters in text-based CAPTCHAs for intelligent character recognition

Over last few years, CAPTCHAs are ubiquitously found on internet as a security mechanism to distinguish between humans and spams. The text-based CAPTCHAs offer users to recognize the distorted text from the challenged images. Having based on hard AI problem, they have emerged as a hot research topic in computer vision and machine learning. The contemporary text-based CAPTCHAs are based on the segmentation problem that involves their decomposition into sub-images of individual characters. This is a challenging task for current OCR programs which is not yet solved to a great extent. In this paper, we present a novel segmentation and recognition method which uses simple image processing techniques including thresholding, thinning and pixel count methods along with an artificial neural network for text-based CAPTCHAs. We attack the popular CCT (Crowded Characters Together) based CAPTCHAs and compare our results with other schemes. As overall, our system achieves an overall precision of 51.3, 27.1 and 53.2% for Taobao, MSN and eBay datasets with 1000,500 and 1000 CAPTCHAs respectively. The benefits of this research are twofold: by recognizing text-based CAPTCHAs, we not only explore the weaknesses in the current design but also find a way to segment and recognize the connected characters from images. The proposed algorithm can be used in digitization of ancient books, handwriting recognition and other similar tasks.     

Sentiment classification of online reviews to travel destinations by supervised machine learning approaches

The rapid growth in Internet applications in tourism has lead to an enormous amount of personal reviews for travel-related information on the Web. These reviews can appear in different forms like BBS, blogs, Wiki or forum websites. More importantly, the information in these reviews is valuable to both travelers and practitioners for various understanding and planning processes. An intrinsic problem of the overwhelming information on the Internet, however, is information overloading as users are simply unable to read all the available information. Query functions in search engines like Yahoo and Google can help users find some of the reviews that they needed about specific destinations. The returned pages from these search engines are still beyond the visual capacity of humans. In this research, sentiment classification techniques were incorporated into the domain of mining reviews from travel blogs. Specifically, we compared three supervised machine learning algorithms of Naïve Bayes, SVM and the character based N-gram model for sentiment classification of the reviews on travel blogs for seven popular travel destinations in the US and Europe. Empirical findings indicated that the SVM and N-gram approaches outperformed the Naïve Bayes approach, and that when training datasets had a large number of reviews, all three approaches reached accuracies of at least 80%.     

粘连字符的图片验证码识别

验证码在维护互联网安全、防止机器恶意攻击做出了很大贡献。但通过现有的模式识别技术仍然可以破解部分验证码。着重于有粘连字符的猫扑和西祠胡同网站验证码进行识别,难点在于分割图片中的粘连字符。对字符是模糊粘连的猫扑验证码,提出了基于局部极小值和最小投影值的方法来分割;对有交错粘连的西祠胡同验证码,通过颜色聚类与竖直投影结合来达到分割字符的目的。最终均采用卷积神经网络进行训练和识别,达到了较高的识别率。     

CAPTCHA Security: A Case Study

CAPTCHAs have been widely used across the Internet to defend against undesirable or malicious bot programs. In this article, the authors describe the security of a CAPTCHA reported in a recent peer-reviewed paper and deployed on the Internet. They show that although this scheme was effectively resistant to one of the best optical character recognition programs on the market, they could break it with a success rate of higher than 90 percent by using a simple but novel attack. In contrast to early work that relied on sophisticated computer vision or machine learning algorithms, they used simple pattern recognition algorithms that exploited fatal design errors. The main contribution of their work is that simply counting the pixels in a CAPTCHA's characters can be a very powerful attack.     

基于优化支持向量机的人脸表情分类

分析了支持向量机（support vector machine, SVM）目前主要存在的问题和参数选择对分类性能的影响后, 提出了以改进粒子群算法优化SVM关键参数的优化SVM算法。将加入拥挤度因子的微粒群算法引入到SVM中, 在不牺牲泛化性能的前提下, 对其参数进行优化, 增加了SVM初始化参数的多样性, 减慢了局部搜索, 促进其在全局范围内的寻优搜索, 以有效克服SVM算法过分依赖初始值和容易陷入局部极小值的缺点, 并利用由粗到精的策略构造多层SVM人脸表情分类器, 在提高准确率的基础上加快分类的速度。实验证明, 新算法具有速度快、准确率高的优点。     

基于边缘与SVM的车牌自动定位与提取

提出了一种将边缘与SVM相结合的车牌定位与提取的方法。首先根据字符的边界特征进行粗筛选,获得几个车牌候选区;然后使用SVM分类器进行字符与非字符分类;最后根据车牌特征实现定位与提取。实验表明,该方法取得了良好的效果。     

卷积神经网络识别汉字验证码

验证码今已广泛应用在各个领域,常见的英文字母与数字组合的验证码自动识别准确率已达到较高的水准,而汉字因其字符复杂,用传统方法进行自动识别难度很大。提出一种基于卷积神经网络的验证码自动识别方法来提高字符的识别准确率。采用Keras卷积神经网络框架,设计多层卷积来提取深层次图像信息,分别对汉字验证码和字母数字验证码进行识别,以提高模型的泛化性。实验结果表明用该方法汉字验证码的单字识别率已达到99.4%;传统四字符字母数字验证码的识别率最高达到99.3%。这一结果表明深度神经网络对验证码复杂结构的感知能力很强大,通过对比实验发现Keras框架在验证码识别领域有较好效果。