ReCAPTCHA

The CAPTCHAs can be more or less difficult but over everything they are very annoying by the distrust and the loss of time that suppose.

In the commentaries of this we used blog them like bad minor to avoid the indiscriminate Spam and the truth is that they work reasonably well. For a computer it is relatively difficult to recognize letters that sufficiently are deformed. For a normal human the process usually is the sea of simple.

Leaving that surely are better solutions the truth separate it is than it is difficult to avoid the CAPTCHAS use at the present time so, since the distrust is going to follow there, perhaps we pruned to improve the lost subject “of time”.

A VERY GOOD IDEA™ is the one that they have had in ReCAPTCHA. One is to take advantage of the CAPTCHAS to digitize texts correctly. That is to say, a voluntary and distributed species of OCR. In order to locate a little to them I will say to them that the Optical monitoring systems of Caracteres (OCR) are used to failing more than a fair gun and at the time of digitizing always printed texts it is necessary to review the turn out to make sure that to the OCR program him the clamp has not gone away.

Then, the idea of ReCAPTCHA is very simple, whenever querramos to validate to us like humanos™ in a Web that uses the ReCAPTCHA system appears to us an image with two scanned words and we do by rough estimate of OCR, that is to say, digitize the words. Thus we will help to digitize whole books with our small contributions.

If, what they are thinking: if we are those that “we translate” the word correctly and their programs of OCR cannot do it then how knows the system we have done that it well and therefore we are human? Then very simple, they put two words to you, one of them it is of control, it was recognized correctly deformed at the time (by its program of OCR) and later, whereas the other is the word who really we must recognize because its program cannot. As we do not know what is what we must write the two well (that avoids possible gamberradas) and since there is to correctly write at least one of the two words cannot be done by programs of automatic way.

A shining and tremendously useful solution.

Via: | ReCAPTCHA Helping to digitize old books.