International Journal of Research and Engineering Issue 2, Volume 1 28 http://www.ijre.org ISSN 2348-7852 (Print) | ISSN 2348-7860 (Online) Hand Gesture Recognition Based 3D CAPTCHA Mr.Vaibhav Kanth Sinhgad College Of Engg. lonavala,Maharashtra vaibhav.kanth@gmail.com Mr.Vivek Dubey Sinhgad College Of Engg. lonavala,Maharashtra vdmedia.co.in@gmail.com Mr. Amit Kumar Sinhgad College Of Engg. lonavala,Maharashtra amit.destini@gmail.com ABSTRACT-CAPTCHA stands for Computer Automated Public Turing Test to tell Computers and Humans Apart which is a widely used security technique to avoid spams and automatic form submissions by bots, thus ensuring that the end user is a human. An alternative to the present Text Based CAPTCHA which is based on Hand Gesture Recognition and Pattern Matching is proposed in this paper. This approach will verify user as human and prevent bots from spamming applications and services thus making the entire process more secure and reliable. This technique will generate a random series of number which the user will be prompted to mimic as gestures with their hands and it will be captured by the Computer Webcam or mobile phone. A pattern matching algorithm based on SIFT algorithm will run on the user provided Images to identify maximum key- points after comparing from the database of the pre-stored standard gesture images and after finding a successful match, the site will be navigated to the next page. This process is highly efficient as it is very difficult to design a bot to identify gestures in images. Key words: Hand detection, hand posture recognition, feature extraction,SIFT, CAPTCHA 1. Introduction CAPTCHA (Completely Automated Turing test to tell Computers and Humans Apart) is widely used security mechanism on the web to ensure that response is from a person and not a bot.Hand Gestures provide a natural and intuitive communication modality for human–computer interaction. Efficient human computer interfaces (HCIs) have to be developed to allow computers to visually recognize in real time hand gestures.An image of misrepresented letters is dynamically generated. The letters are part of image not a plain text CAPTCHA which uses a type of challenge response test to determine that the response is not generated by the computer. Now a day’s all sorts of websites are using small to large CAPTCHA’s [1]. Fig. 1 Commonly used CAPTCHA consists text based challenge is given to web clients to be typed in textbox where humans can pass the challenge but not the machine. For example a human can easily read misrepresented text shown in figure 1 but current programs cannot. CAPTCHA is mainly used to avoid spam.Spamming is performed on various public email provider sites and also on various forums and blogs. This comes into picture where users need to submit some content to a service or web applications.Even though many types of CAPTCHA are implemented they fail at some time, for e.g., background CAPTCHA are easily hacked by using simple computer visiontechniques [1]. CAPTCHA should be in such a way that it should be easily understandable by humans and should be easy to answer or interact with the process, it should be very hard forbots or machines to solve or it should be only understandable by humans, while bots cannot understand the process. And also there should be an easy process to generate and evaluate CAPTCHA and should not produce network overheads. Designing a CAPTCHA which satisfy these requirements is not so easy and that’s the reason some times CAPTCHA [1] fail. Keeping in mind these systematic rules, our system performs 90% accurate process of identifying human. We generate simple clear images of characters that are not blended or twisted so it is really understandable by humans. Now we provide a new way of user interaction with CAPTCHA process, that is acquiringanswer via image of user gesture either using webcam or mobile phone camera. Using this way bots cannot submit image, only user can submit gesture image [8, 9, 10.]. We perform a pattern matching algorithm and verify if user submitted image is a gesture of particular character [5, 6].Our system is highly intuitive and easily understandable by humans and user feel it more appealing than annoying traditional image based CAPTCHA [1].However, vision-based hand tracking and gesturerecognition is a challenging problem due to the complexity ofhand gestures, which are rich in diversities due to high degreesof freedom (DOF) involved by the human hand. In orderto successfully fulfill their role, the hand gesture HCIs haveto meet the requirements in terms of real-time performance,recognition accuracy, and robustness against transformationsand cluttered background. 2. Working of CAPTCHA Many types of CAPTCHA are developed and implemented to prove particular subject is human or any kind of programmed bot. There have been various attempts at creating more accessible CAPTCHAs, including the use of JavaScript, mathematical questionshe most widely used CAPTCHAs are the so-called text- based schemes, which rely on sophisticated distortion of text images aimed at rendering them recognizable to the state of the art of pattern recognition programs [5].The popularity of such schemes is due to the fact that they have many advantages, for example, being intuitive to users world-wide (the user task performed being just character recognition), having little