Implementation of Haar Cascade Classifier for Vehicle Security System Based on Face Authentication Using Wireless Networks

This research work presents a vehicle security system for safeguarding the vehicle from theft issues under the architectural design of capturing and comparing the vehicle user's face. Since the face of human beings are unique and has different biometric characteristics which are complex to make fraudulent activities. Authenticating the vehicle users with face recognition mechanism is highly secured than token-based and knowledge-based security mechanisms. This research ultimately models and classifies the vehicle users into authorized and unauthorized users. Initially, an experimental prototype for vehicle security system is developed, and the application of image processing algorithms is incorporated into the model. The system uses Haar feature-based cascade classifier and AdaBoost method which is a machine learning algorithm used for detecting the authorized user's face effectively. The algorithm is trained initially with appropriate amount of positive and negative images, and the feature gets extracted. When the person tries to access the vehicle, the experimental system captures the image of the person and makes comparison with extracted features to identify the authorized user. Finally, the results obtained from the prototype system are satisfied and beneficial against the issue of vehicle theft.