Meanwhile I had some email conversation with UPEK. They pointed me to their new BSAPI_3.0_for_Linux available for download here:
http://www.upek.com/solutions/pc_and_networking/sdks/linux/ I've tested the samples and they worked. The API doc looks complete. They provide a (closed source) libbsapi.so that is for x86 systems only. This is what they say about: --------- BSAPI: - Single library, no need to install BioAPI - Supported devices: TCD41, TCD42, TCS4B, TCS4C, TCS4E, TCD50, TFM 2.0 - Biometric functions are done by host, so Sensor only solutions are supported as well. Basically all current UPEK devices are supported except plain area sensor reader (TCS1 and TCS2). - Authentication module for PAM is NOT included! This SDK is designed more for Linux development. - BSAPI offers much wider range of functions than BioAPI BSP - Only x86 binary is available --------- ...and later on: --------- - Identification can be easily done by calling VerifyMatch function in loop. This function is available in both SDKs, just the input/output parameters are slightly different. Basically if you have 1000 templates in your database, you need to call VerifyMatch function 1000 times (once for each template in database) to identify the right user/template. As matching is done on host, it is very, very fast, no worries about performance! --------- They have denied to make their libbsapi open source. I don't know how you guys think about that, but my recommendation would be to "wrap" something around this libbsapi.so so it can fit into libfprint as a driver with full capabilities of the UPEC devices. It would then be up to the user to install or not the libbsapi.so in their systems. The same has been seen with ATI display drivers. Expecting your violent protest ;-) - Regards Wolfgang Ullrich e-Mail: [EMAIL PROTECTED] Network administrators first choice: N-View, THE network monitor! http://www.n-view.net More good software wanted? Look here: http://www.pdfserver.net (A Network PDF Server Appliance) Am Mittwoch, den 13.08.2008, 12:24 +0100 schrieb Richard Lloyd: > We've got this scenario we want to develop fingerprint reading code for: > > * Multiple low-end PCs, each with a USB fingerprint reader attached. > For ease of development, all PCs will be the same hardware, run the same > Linux and run the same fingerprint driver/app software. Similarly, all > fingerprint readers will be identical hardware. PCs will be such a low > spec that software image comparison is probably not advisable, but we > could go that route if there's no hardware comparison is available. > > * The number of people using the fingerprint readers across the PCs > could rise as high as 100 or more and it's likely we'd take at least > two prints (e.g. thumb and a finger) per person. > > * Enrolment should be available on *any* reader and only need to be done > once overall (i.e. not on every PC/reader), with the image/template > retrieved distributable to the other PCs (i.e. the software driver or > the fingerprint reader hardware needs to do one-to-many comparison, > potentially for up to 200 prints in one go and have enrolments > from other units "uploadable" to it). > > With low spec PCs, I suspect comparing images of a scanned print with > up to 200 other images would take a long time (10 seconds or more?) - > too long for someone waiting to be verified, so I feel a hardware > comparison solution would be the best. > > The UPEK Eikon reader looked promising, but seems to fail on a couple > of points: > > * The Linux driver only does one-to-one comparison and even the > Windows driver claims to only do a max of one-to-21. > > * There is no clear indication whether the unit can be programmed > with previous enrolment templates (200 bytes of data I believe) > taken from another Eikon reader. From what the page at > http://reactivated.net/fprint/wiki/Upekts suggests, it seems that > you have to put the unit into verification mode, manually swipe > and it just does a 1-to-1 comparison with a single print stored > in the unit (with no clue as to whether the only way to get that > stored print into the unit is to enrol it manually or not). > I could be wrong on this point though... > > I would have thought that in a small/medium business that has multiple > exits/entrances to a building (we want to do this for a clock in/out > system, rather than a PC login or door access system) and wants to > use fingerprint readers at all exits/entrances, then the above > scenario would be quite common. For cost reasons, we want to keep > each fingerprint reader under $200 and don't want to pay for expensive > SDKs either, but my gut feeling tells me that only "industrial" > fingerprint readers will meet our requirements. > > Ideally, we'd like to look at hardware supported by the fprint project > first, because we'll be coding it in Linux (Windows is not an option before > you ask). Does anyone know of a sub-$200 USB fingerprint reader that has > Linux support and fits our criteria? I've been Googling around without > much luck so far :-( > > Richard K. Lloyd, E-mail: [EMAIL PROTECTED] > Connect Internet Solutions, WWW: http://www.connectinternetsolutions.com/ > First Floor, Faraday House, > Liverpool Digital, 360 Edge Lane, > Liverpool, > Merseyside, UK. L7 9NJ > > > > -- > This e-mail (and any attachments) is private and confidential. If you have > received it in error, please notify the sender immediately and delete it > from your system. Do not use, copy or disclose the information in any way > nor act in reliance on it. > > Any views expressed in this message are those of the individual sender, > except where the sender specifically states them to be the views of Connect > Internet Solutions Ltd. This e-mail and any attachments are believed to be > virus free but it is the recipient's responsibility to ensure that they are. > > Connect Internet Solutions Ltd > (A company registered in England No: 04424350) > Registered Office: Faraday House, Liverpool Digital, Edge Lane, > Liverpool, L7 9NJ > Telephone: +44 (0) 151 282 4321 > Fax: +44 (0) 151 282 4322 > VAT registration number: 758 2838 85 > _______________________________________________ > fprint mailing list > [email protected] > http://lists.reactivated.net/mailman/listinfo/fprint > _______________________________________________ fprint mailing list [email protected] http://lists.reactivated.net/mailman/listinfo/fprint
