Signature Verification NIT B.tech Final Year Computer Science Project Report

In office automation recognition of software and document analysis is highly required. To process handwritten small samples with efficiency great ability is required and it includes those that are found on envelopes, checks and so on. In the current research this is the major driving force. There is a large variation between the handwriting samples that is creating a great hamper. Researchers are now carrying on with their research work to find various techniques that can improve the computer’s ability to recognize as well as representing handwritten samples.

The use of artificial neural network has shown a great impact in one of the approaches. In this approach an artificial neural network is used that can trained to locate similarities and patterns between various written samples. In this project we will explore the topic of automated signature with the help of artificial neural network.

This new system is designed to recognize the signature automatically from a JPEG image. The recognition part is based on the neural network and enhancement can be made to the system by making modifications in the algorithm weight adjustments. There are certain limitations that prevailed in this system but I believe that had given my best to make this system efficient and more meaningful. The same neural network can be used to recognize the applications of the pattern.

Only a slight modification can be required. This kind of system is quite useful and it is helpful for big  MNC companies and even small business and organizations can  take its advantages. Nowadays frauds have to a high extent and people can easily copy signatures and make malicious use of it but with the introduction of this system it can no longer be done. Moreover, this system promises to increase the efficiency and communication system. This system can make work easier to handle.