I was looking for an SDK to learn a bit about fingerprint identification system. I have googled a lot and found many different SDKs. However none of them were free. I dont want to implement any commercial system, just want to work with an SDK for educational purpose. Do you guys know of any free biometric SDK available for C or C++?

I guess I found this:
GrFinger Biometric SDK

Personally, I don't see how that will help you especially if your database is meant for educational purposes.

If you're just going to import the entire thing into your project what does it show your teacher/lecturer:- that you know how to import third party libraries? Wow, impressive!

I'm not saying fingerprint id is easy to do from scratch either. Far from it. The algorithms for today's systems are advanced.

You'll have to simplify it right down. Maybe to the point where you draw your own sample fingerprints, with one color and one pencil thickness in paint or something. Then you would apply a simple AI algorithm to try and match the current fingerprint with one on your database.


In any case, I would assume the key purpose of this project is to test your knowledge of databasing.

I would think about organising your records into an efficient heirarchial structure much like a binary search tree. Then you can go about efficiently searching your database. These are probably the things that will most impress your lecturer/teacher.


Do the fingerprint stuff after. If you manage to complete it -great, think of it as a bonus rather than a necessary feature.

Me thinks. Tee he he

My database is not for commercial use. I needed an SDK which was free for 'non-commercial' purpose. I preferred to use 'educational' in this context.

Ha ha. The funny thing is that it will be just enough to impress my lecturer.
I am planning to implement the database in Oracle. Communicating and intergrating with the Oracle database(through ODBC?) is going to be very educational for me as it is one of the important parts of our database course.

It is not possible for me to develop my own algorithm on fingerprint id at this moment. I simply have no time to spend on this. I would love to since this topic really intrigues me. But I do not even have the rudimentary knowledge on "fingerprints" and I will have to spend hours(which I don't have) researching before I can attempt anything. For the moment I just need a function that takes two images of fingerprints and returns whether they match or not. I dont even care about threshold values, simple boolean true/false would work just fine. I dont even mind using the same function for identification as well as for verification.

YES.

Efficiency is the last thing I have in my mind right now. Even if the algorithm takes 10 seconds to match two fingerprints I have no problem with it. I first have to make a skeleton of the database. If time permits only then I can try to make things efficient.

I am. That's why I somehow need to integrate it with my database(which is still far from being implemented), and I do not care much about whether using SDKs or by hardcoding myself. Because my lecturer is not going see whether I did it myself or used a third-party library.

With my experience at my current Univerisity the lecturer will be impressed only based on the front-end of my database. Even if I make the best back-end but a poor front-end I wont get a good grade. On the other hand, even if my backend is poorly made but I have a nice attractive front end with loads of features that will definitely kick ass. The lecturer has little time to delve into what's inside. In my many previous projects I made a lot of useful programs in C++ in console, but I got less marks compared to my friends who made stupid programs with poor performance but had a nice GUI in Java.

For this reason for the last two semesters I have changed my approach. I now try to show off as much as possible so that girls in my batch also say WOW when they see my project. I know it's bad and I will miss the opportunity to learn things that are more important. But sometimes when you don't get credit for your hard work it gets tough.

>The lecturer has little time to delve into what's inside. In my many previous projects I made a lot of useful programs in C++ in console, but I got less marks compared to my friends who made stupid programs with poor performance but had a nice GUI in Java.

This makes me laugh. Tee he he. This also...

>For this reason for the last two semesters I have changed my approach. I now try to show off as much as possible so that girls in my batch also say WOW when they see my project.


I've never done databasing before. But it sounds as if all your lecturer is concerned about is you having something that looks good and has good functionality.

How you did it, or the code behind it is not important. I can't see how that might be beneficial but hey, like I said I've never done DB before kiddo. :surprised

True.
In the long term it is not, except for the students who don't know much about programming except for making good GUIs in Java.

yes i also working on it. and i do some work on it. but i could not work on device. i only work on matching of Fingerprint. but Capturing the image from device. i am thinking on it. please tell me u can only capture routine which is interact with the Fingerprint Device driver. if u have any work on it then we can share.

Last post was March 15th, 2006. In the before posting thread it says don't post on dead threads.

I Know that this is very much old thread. But from ur statment it's sound u got the solution of this old problem thread. if it is right then please tell me.