EXTENDED MOBILE CAPTCHA IMPLEMENTATION ABSTRACT: A CAPTCHA is a type of challenge-response test used in computing as an attempt to ensure that the

response is generated by a person. The process usually involves one computer (a server) asking a user to complete a simple test hich the computer is able to generate and grade. !ecause other computers are assumed to be unable to solve the CAPTCHA" any user entering a correct solution is presumed to be human. Thus" it is sometimes described as a reverse Turing test" because it is administered by a machine and targeted at a human" in contrast to the standard Turing test that is typically administered by a human and targeted at a machine. A common type of CAPTCHA re#uires the user to type letters or digits from a distorted image that appears on the screen.$ost interactive sites today are run by databases and become #uickly clogged and sluggish operating server can handle. &tandard CAPTCHA mechanisms don't ork ell on the mobile eb. The goal of the $obile CAPTCHA pro(ect is to embrace the limitations of mobile platforms to create a usable CAPTCHA mechanism that is also secure against bots. It is comparatively rite an automated script to discover <form> elements ithin the source code. )e follo ed the *+,-'s !rainstorming rules to come up ith several potential designs for a ar!er to a"tomate !ata s"#missio$ %it i$ $ative apps& This is due to the fact that you cannot (ust hen a database table e%ceeds capabilities the

mobile-friendly CAPTCHA mechanism" including arranging image fragments" shaking the phone in different patterns" a .)heel of /ortune. interface" and a social net ork face recognition. )e observed users performing tasks here one of the steps as solving a CAPTCHA on a

mobile phone" like signing up for a 0mail account or purchasing event tickets on ticketmaster.com. )e came up ith three design sketches based on these results. The design used for our video prototype combines t o e%isting image-recognition CAPTCHA mechanisms into a mobile form factor. /irst" users need to identify images from a specific category" then they need to trace an outline of the image on their touchscreen.

EXISTIN' S(STEM: &tandard CAPTCHA mechanisms don't ork ell on the mobile eb. The goal of the $obile CAPTCHA pro(ect is to embrace the limitations of mobile platforms to create a usable CAPTCHA mechanism that is also secure against bots. It is comparatively rite an automated script to discover <form> elements ithin the source code. DISAD)ANTA'E: Are not compatible ith users ith disablilities Time-consuming to decipher Technical difficulties ith certain internet bro sers $ay greatly enhance Artificial *ntelligence ar!er to a"tomate !ata s"#missio$ %it i$ $ative apps& This is due to the fact that you cannot (ust

PROPOSED S(STEM: An online test that humans but not computers are able to pass" used as a security measure and usually involving a visual-perception task1 &ite visitors must solve the 2distorted te%t3 CAPTCHA before posting comments. A type of test used to determine hether a re#uest to a eb site comes from a human or a computer program" typically by asking the user to perform some kind of image recognition task such as reading distorted te%t. The term as coined in 4555 by 6uis von Ahn" $anuel !lum" 7icholas 8. Hopper (all of Carnegie $ellon 9niversity) and 8ohn 6angford (of *!$) as a contrived acronym for .Completely Automated Public Turing test to tell Computers and Humans Apart.. CAPTCHA aims to prevent soft are tools from performing actions hich might degrade the service" such as registering user accounts or automating the playing of a game. AD)ANTA'E: +istinguishes bet een a human and a machine $akes online polls more legitimate :educes spam and viruses $akes online shopping safer +iminishes abuse of free email account services

CAPTCHAs are used in attempts to prevent automated soft are from performing actions hich degrade the #uality of service of a given system" hether due to abuse or resource e%penditure.

)hen a user starts a registration process" the backend server starts a parallel process on the actual target ebsite. )hen a Captcha is re#uired by the target ebsite" the image is then transferred and

displayed as part of the 2free3 ebsite registration process. Any parsing provided by the real user is then transferred as input to the target ebsite.

MOD*LES: +& Pre,processi$-: Removal o. #ac/-ro"$! cl"tter a$! $oise& /irst preprocessing is needed to reduce the clutter to a level here k-means clustering can be applied. CAPTCHAs do not have to rely on difficult problems in artificial intelligence" although they can. &urely you;ve encountered today;s ridiculous image Captcha routines" here they scramble a bunch of letters in a funny looking image" hich is and then ask you < the human < to decipher and enter the te%t in a bo%"

checked for a match before submission. +eveloping a CAPTCHA application isn;t tremendously difficult" but it does involve te%t-graphics manipulation and re#uires your server to remember 0& Se-me$tatio$: &plitting the image into regions hich each contain a single character. The only step here humans still outperform computers is segmentation. *f the background clutter consists of shapes similar to letter shapes" and the letters are connected by this clutter" the segmentation becomes nearly impossible ith current soft are. Hence" an effective CAPTCHA should focus on the segmentation. Traditionally" defeating a CAPTCHA test re#uires t o procedures1 segmentation and recognition. :ecent research sho s that the problem of segmentation is much harder than recognition. *n this paper" t o ne and character arping. 1& Classi.icatio$: *dentifying the character in each region. /or classification of the segmented numerals obtained" e have used support vector machines using individual pi%els as feature e find the best fit rectangle for the numeral vectors. After segmenting the numbers" segmentation techni#ues called projection ith line cluttering and middle-axis point separation are proposed for CAPTCHAs hat the CAPTCHA is for your visitors session so they can match up the response ith the original string.

e have done vertical segmentation using histogram.

and then scale it to a prefi%ed resolution. The toughness of the separating numeral from clutter increases ith the number of background lines. ,ach pi%el of the image is analysed and its $anhattan distance from its neighbouring =-pi%els is computed. The binary-image filter aims at using the vacant space present around individual pi%els to remove secluded pi%els from the image obtained after color-based filtering. The resultant image from binary filter has many small connected components of clutters. 7ote that the pi%el count of most of these noise patches is significantly less than the pi%el count of the numerals.

S(STEM RE2*IREMENTS: HARD3ARE RE2*IREMENTS: &ystem Hard +isk /loppy +rive $onitor $ouse :am 1 1 1 1 1 1 Pentium *> 4.? 0H@. ?5 0!. A.?? $b. AB >0A Colour. 6ogitech. BA4 $b.

SO4T3ARE RE2*IREMENTS: -perating system Coding 6anguage Tool Eit *+, 1 1 1 1 )indo s CP. 8ava A.D Android 4.4 ,clipse