Back Propagation ANN algorithm and Database?

hello everyone!

I'm doing a project "Signature Recognition and Classification System" I use the Zernike moments for feature extraction and for classification I use Back Propagation Artificial Neural Network, usually every signature is related to a person so, When the signature assign in the system, the name, last name and ID of person with the feature of signature is stored in a database (I use mat file for storing data), My question is that how to store this data in mat file that later I could Identify signature and compare new signature with database signature, Can I store this data in dataset? Can I store in structure array? and so on. What do you suggest and advise me? if you want get more information about my project see below link:

http://www.irjcjournals.org/ijieasr/jan2013/1.pdf

thanks a lot. please help me I need your suggestion very much.

Mansoor

[Mansoor ahmadi]

please answer my above question, I need it hard.

[Stephen23]

You should change the name of the question from "Need to help in my project !!!!!!" to something more descriptive. If you want people who are browsing MATLAB Answers to come and help you, then you need to help them by giving them the information so they know what topic you are working with.

http://www.mathworks.com/matlabcentral/answers/6200-tutorial-how-to-ask-a-question-on-answers-and-get-a-fast-answer

[Mansoor ahmadi]

Thank you sir for your advise!

I changed my question, I'm new in this site I don't know some this website's rule. Now if you can answer me. thank you.

Answer 1

[Stalin Samuel]

for example(basic &not NN) if you have featire f1 f2 f3

   f(i,:)=[f1(i) f2(i) f3(i)]
   save feature
then use 

   load feature
and calculate feature for a single image and match with the data base