CAPTCHA, those distorted letters people often encounter when filling in web forms, can be found on an increasing number of web sites even though the can be annoying for users. Here is a look at what they are, why they are used and the arms race between those who want to keep spammers and others from web sites and those who want to misuse the internet.
CAPTCHA is an acronym standing for Completely Automated Public Turing Test to Tell Computers and Humans Apart. As the name indicates, they are used to verify that a person, and not a software program, is behind the submission.
Since humans are better at certain tasks than computers, CAPTCHA are used in situations where spammers and others with malicious intent may use software to abuse web sites. For example, spammers use programs to automatically submit links promoting spam products through web forms. Or, they may use these programs to automatically create email accounts on respected providers such as Gmail and then use those accounts to send spam.
CAPTCHA often consist of distorted letters set in backgrounds that make them more difficult to read. Since the human brain is still much better at recognizing patterns in these types of situations than computers, a properly designed CAPTCHA can still defeat computer programs most of the time. However, spammers and others are developing increasing sophisticated ways to defeat circumvent them.
One way of doing this is by using OCR (turning image text into a format that can be read or edited by a computer). This is usually done in three stages:
1. First, the background clutter is removed as best as possible. For example, different colors, put onto the background of the image to make it more difficult for a computer to read, can by circumvented by having the program first remove colors from the image.
2. Then, the image is divided into regions with each region containing one character
3. Finally, the program can then identify and classify each image.
While computers do not usually have much trouble with step one and three, segmentation is still difficult for computers when the images are connected by background clutter. Thus, images that are most difficult to segment work the best.
There are other methods of circumvention, often used in connection with OCR, to include:
1. Deciphering all the CAPTCHA images (when there is a limited pool) and applying them using tables
2. If the program is incorrectly designed and allows more than one attempt at the same image, it can be bypassed when the image becomes known.
3. Cracking the hash being used with the CAPTCHA when the image is too small
A relatively low-tech method involves sending the images to workers in low-wage countries to decipher in bulk.
As the methods and algorithms used by those trying to circumvent the system become more sophisticated, CAPTCHA have become more difficult. While new ways using pictures, short reading comprehension and other methods are being developed, none has proven to be foolproof. There is also the danger of turning legitimate users off if the process becomes to onerous.
The one silver lining is this arms race is that it deals with making machines think like people and is thus strengthening the drive to create artificial intelligence.
Regardless, as long as there are people using automated programs to use the internet for malicious intent, CAPTCHA use and sophistication is likely to grow.
Buy Shower Curtains is a comprehensive information source on buying shower curtains, shower curtain care and everything else about shower curtains.
0 comments:
Post a Comment