Hello, I figured it out! The result is on [1]. The image is clearly contained of stripes, which should be stitched together (I hope, there is yet written algorithm in Libfprint, there also might be information about the shift in extensive metadata (see [2]) of the stripes. The noisy strips at bottom are caused by bug in my code. But I am expect, the quality of the images will still not be as good as Алексей's [3].
After I tune up the communication details and perform successful initialization, I could start with writing of the actual driver. I would like to hear some advice from experienced Libfingerprint developers here. I started to write some protocol description [4]. Although it is still working version, you might have a look at it, if you didn't recognize it to be similar to protocols of some other devices, with which it could share driver. You can also have look at the prototype source, if you are interested [5]. [1] http://cozpserver2.jinonice.cuni.cz/~petr/download/fingerprint/fingerprint.pgm [2] http://cozpserver2.jinonice.cuni.cz/~petr/download/fingerprint/fingerprint-fibre.pgm [3] http://gentoo.homelan.lg.ua/distrib/final-enrolled.pgm [4] http://cozpserver2.jinonice.cuni.cz/~petr/download/fingerprint/codes.txt [5] http://cozpserver2.jinonice.cuni.cz/~petr/download/fingerprint/aes1660-prototype.tar.bz2 Dne Mon, 21 Mar 2011 15:04:27 +0100 Petr Dlouhý <petr.dlo...@email.cz> napsal(a): > Hello, > > by good luck, I found out, the way, how to transmit the data from the > fingerprint reader unencrypted. I was able assemble a file, which has > clear bitmap-like data - see [1, 2]. > > When you open the file in Gimp or other bitmap editor, you can clearly > see > header (to byte 43) and footers (from byte 555) with some kind of > additional data (that might be also interesting), between them is 512 > bytes of the fingerprint bitmap. That is, how it appears to me. > > Although, I was not able to make a result with fingerprint clearly > visible. > > What is your opinion about this file? Do you think, that it is a real > fingerprint data? How is it encoded? > > [1] http://www.uloz.to/8370560/data-unencrypted1-pgm > [2] http://www.uloz.to/8370757/data-uncrypted-data-clearly-visible-pgm > > Dne Mon, 21 Mar 2011 00:34:30 +0100 Petr Dlouhý <petr.dlo...@email.cz> > napsal(a): > >> Hi, >> Now I realized some new facts about the fingerprint data. The header >> "4954 >> 020d" is repeated (even longer sequence is repeated) every 1196 bytes in >> the stream. >> This could actually mean, that the transfer in fact consist of strips. I >> tried to assemble PGM image from the strip, but I didn't find any >> structure in the data. >> Any suggestions? > > -- Petr Dlouhý _______________________________________________ fprint mailing list fprint@reactivated.net http://lists.reactivated.net/mailman/listinfo/fprint
