Person Verification by Behavioral Bio-metric System

Research focusing on identification of a person using behavioral biometrics till now has only been done using exorbitant hardware like pen, tablets or sensors for capturing the motion of wrist. In this project, we are developing a framework for real vs fake signature identification by acquiring data

Project Title

Person Verification by Behavioral Bio-metric System

Project Area of Specialization

Wearables and Implantable

Project Summary

Research focusing on identification of a person using behavioral biometrics till now has only been done using exorbitant hardware like pen, tablets or sensors for capturing the motion of wrist. In this project, we are developing a framework for real vs fake signature identification by acquiring data using wrist worn devices. A smart sensor-based framework that is capable of recognizing gait (walk) and signature pattern of a user will be developed in this project. Motion corresponding to the gait and signature pattern of a user will be captured by the smart sensors. Machine learning classifiers will be trained on a dataset of gait and signature pattern acquired using smart sensors and tested by using the test data. This acquired hand waving and signature pattern will be matched with a database to authenticate the user and mark his or her identity.

Project Objectives

The aim of the project is to develop an attendance system based on behavioral biometrics.

The objectives set are:

1. Understanding the Smart sensor signals.
2. Data collection.
3. Applying machine learning algorithms and analyzing the results for identifying the identity and real vs fake signature verification.
4. Real time implementation.

Project Implementation Method

The scheme is composed of data acquisition, feature extraction, feature selection, and signature (genuine and fake) identi?cation. In data acquisition we collect data of sensor of student’s signature in normal condition. A set of sixteen-time domain features are extracted from the acquired sensors data. The set of extracted features have earlier been used for recognizing signatures of a person. The extracted set of features are passed through different method for feature selection to select the features which are least correlated among each other and are highly correlated with the class labels. Then apply different algorithm is used for the recognition of real and fake signatures using the acquired inertial sensors data. After classifier selection we intend to implement the system in-real time.

Benefits of the Project

The main benefit of our project is that the identification and verification of an individual becomes easy and cheaper which can be used further for security purpose. It can also be used for employee attendance, door security as well as logical access. Some of the advantages of or project is as follows:

• Accountability
• Time saving
• Easy to use
• Accuracy
• User-friendly
• Cost-effective

Technical Details of Final Deliverable

In this project, we intend to develop a mechanism for marking the identification of the employees or students based on their hand waving and signature patterns using smart sensors.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Legal

Other Industries

Core Technology

Wearables and Implantables

Other Technologies

Sustainable Development Goals

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com