Now this method is a little bit harder. But what are the advantage if you can do this hard part then what are the advantages ?
According to me the advantage is tremendous.
You can have your own web shop/business etc shop for selling the authentication devices, attendance systems etc.
But what are the hard parts ?
What are the price ?
Hard parts are that you need to be expert with microcontroller and interfacing and need to be a good student with a capability to read and learn almost any circuit ,computer programming etc terms by yourself without anybody’s help, because at this level almost nobody will help you as this may be an industrial product and when the term comes money then nobody will give you the code etc and t even not tell anything about that.
I am here giving a rough idea about what you need and how to do that..
You need to have a camera ( not web cam rather small standalone camera chips or image sensor chips ) then image sensor chips driver need to be written for microcontroller. May be you need to go for a better microcontroller like PowerPc or ARM etc. Then after that on that microcontroller you need to implement the finger print algorithm .
So this is the way. The price of single camera chip not so low, may be more that $30.
Then you need to learn the microcontroller because some advanced microcontroller needed in this case.
Is there any other way ?
Instead of microcontroller you can use DSP chips may be TMS 6711 chips by Texas instruments and write a camera driver in code composer studio and also implement the algorithm for finger print in that TMS series DSP chips.
FPGA based approach:
In this approach of the design we need to learn Verilog language very well. Then apply the hardware knowledge to design the finger print recognizer engine. Then implement the algorithm using verilog compiler and the use a CPLD/FPGA to download the code to.You may use Xilinx ISE webpack to download your code to the FPGA kits manufactured Xilinx corporation. Now obviously you need to write the camera driver if you use standalone camera for that.
Other Links for same project implementation